review article hereditary fructose intolerancehereditary fructose intolerance (hfi, omim 22960) is a...

ItMed Genet 1998;35:353-365

Review article

Hereditary fructose intolerance

Manir Ali, Peter Rellos, Timothy M Cox

AbstractHereditary fructose intolerance (HFI,OMIM 22960), caused by catalytic defi-ciency of aldolase B (fructose-1,6-bisphosphate aldolase, EC 4.1.2.13), is arecessively inherited condition in whichaffected homozygotes develop hypoglycae-mic and severe abdominal symptoms aftertaking foods containing fructose and cog-nate sugars. Continued ingestion of nox-ious sugars leads to hepatic and renalinjury and growth retardation; parenteraladministration of fructose or sorbitol maybe fatal. Direct detection of a few muta-tions in the human aldolase B gene onchromosome 9q facilitates the geneticdiagnosis of HFI in many symptomaticpatients. The severity of the diseasephenotype appears to be independent ofthe nature of the aldolase B gene muta-tions so far identified. It appears thathitherto there has been little, if any, selec-tion against mutant aldolase B alleles inthe population: in the UK, -1.3% ofneonates harbour one copy of the preva-lent A149P disease allele. The ascendanceofsugar as a major dietary nutrient, espe-cially in western societies, may accountfor the increasing recognition of HFI as anutritional disease and has shown theprevalence of mutant aldolase B genes inthe general population. The severity ofclinical expression correlates well with theimmediate nutritional environment, age,culture, and eating habits of affected sub-jects. Here we review the biochemical,genetic, and molecular basis of humanaldolase B deficiency in HFI, a disorderwhich responds to dietary therapy and inwhich the principal manifestations of dis-ease are thus preventable.(TMed Genet 1 998;35:353-365)

Keywords: fructose intolerance; aldolase B; inborn errorof metabolism

Aldolase B: a specialised enzyme offructose metabolismFructaldolases (EC 4.1.2.13) are widely dis-tributed in living organisms; they catalyse thespecific and reversible cleavage of the glycolytichexose substrates, fructose-1,6-bisphosphateand fructose 1-phosphate, into the 3-carbonsugars, dihydroxyacetone phosphate, D-gly-ceraldehyde 3-phosphate, and D-glyceral-

dehyde' (fig 1). There are two differentprocesses for the generation of the stablecarbanion substrate enzyme complexes in-volved in the aldolase reaction: aldolases of theclass II type, which are confined to yeasts andprokaryotes, form the carbanion using abivalent metal cation (Zn21) .2 However, in classI aldolases, which are mainly of eukaryotic ori-gin, there is transfer of a proton from aconserved Schiff base forming lysine residuelocated at the active site.'There are three genetically distinct isozymes

of class I vertebrate aldolase which can bedistinguished on the basis of their antigenicand catalytic properties.4`6 In humans, aldolaseA (Genebank Accession No Ml 1560) exists inmost tissues but predominates in the muscle;aldolase B, previously known as liver aldolase(Accession No X01098) is also found incertain cells of the kidney and small intestine;and aldolase C (Accession No X07092) ispresent in the brain. Aldolases A and C areconstitutively expressed, but the B isoform isunder dietary control.7 Under resting condi-tions, the expression of aldolase B mRNA islow and any functional enzyme present in thecytosol appears to associate with componentsof the cytoskeletal network rendering it cata-lytically inactive; after feeding a carbohydratediet, mRNA expression is induced and the

GLYCOGEN,---------------.-

FRUCTOSE GLUCOSE

Fructose-1-phosphate Fructose-1, 6-diphosphateI[AldolaseB4----

3 carbon sugars _ Triose phosphates

Pyruvate

LactateFigure 1 The metabolism offructose in normal subjects. Theenzyme aldolase B catalyses the reversible cleavage offructose1-phosphate in the specialised pathway offructosemetabolism, as well as fructose-1,6-bisphosphate in theglycolytic/gluconeogenic pathway. 3-carbon sugars:D-glyceraldehyde 3-phosphate, D-glyceraldehyde,dihydroxyacetone phosphate. Dotted arrows indicate point ofsecondary metabolic arrest in aldolase B deficiency (see text).

University ofCambridge,Department ofMedicine, Level 5,Addenbrooke'sHospital, Hills Road,Cambridge CB2 2QQ,UKM Ali*P RellosT M Cox

Correspondence to:Professor Cox.

*Present address: MolecularMedicine Unit, ClinicalSciences Building, Universityof Leeds, St James'sUniversity Hospital, LeedsLS9 7TF, UK.

353

on February 13, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/

Ali, Rellos, Cox

enzyme is activated by dissociation of the aldo-lase B-cytoskeletal protein complex in thepresence of accumulated intermediatemetabolites8'-3 In fetal tissue, aldolase A is pre-

dominant, but as the embryo develops there isrepression of aldolase A in the liver, kidney, andintestine, accompanied by expression of aldo-lase B.'4`6 A reversion to this fetal pattern ofexpression is observed in transformed rat livercells and in human hepatoma, where aldolase Bhas been categorised as an oncofetalantigen.'7 18

The 5' flanking sequences (up to 200 basepairs) of the aldolase B gene from differentspecies are highly conserved, as expected forcommon promoter regions that are importantfor gene expression. The developmental stagespecific and tissue specific expression ofaldolase B has been shown to be linked with thechromatin structure of its promoter region(which encompasses an origin ofDNA replica-tion initiation),'9 20 the demethylation of a

cytosine residue (position -129),21 and thesimultaneous appearance of transcription fac-tors, AIF-A (identical to liver specific factorHNF-122) and AIF-B (CCAAT binding pro-

tein, binds to residues at positions -129 and-128), which activate aldolase B gene transcrip-tion by binding to the promoter region.23 24 Thetranscription factors HNF-3 and RYB-a,which have overlapping binding sites withAIF-A and AIF-B respectively, have been iden-tified to repress transcription from thispromoter 5 26 suggesting that competition be-tween these activators/repressors regulates al-dolase B gene expression.27 Recently, Sabourinet af8 used transgenic mouse models withexpression construct transgenes to identify an

intronic activator region (nucleotides 650-2448 in the aldolase B gene) which, in combi-nation with the tissue specific promoter, is suf-ficient to mediate developmental expression ofaldolase B mRNA expression but lacks thenecessary site(s) for activation by dietarycarbohydrate.

Studies of enzyme function activity usingpurified isozyme preparations have shown thatat saturating substrate concentrations, aldo-lases A and C display between 50- and 10-foldgreater activity towards fructose-1,6-bisphosphate than to fructose 1-phosphate,whereas aldolase B has equal activity with bothsubstrates.6 29-31 The aldolases show less affinityfor the fructose 1-phosphate substrate (aldo-lase B KM 1 mmol/l), compared with thebisphosphate, (aldolase B KM 50 pLmol/J).6 32 33

The similarity between the isozymes, in terms

of their structural and general catalytic proper-ties, led Rutter34 to conclude that the isozymesevolved by a process of gene duplication anddivergence, according to the mechanism pro-posed by Ingram35 to explain the differencesobserved in the haemoglobins. This view hasbeen supported by the mapping of the threehuman aldolase genes and a pseudogene to

homoeologous chromosomes that haveemerged during vertebrate tetraploidisation.36

Studies on the molecular architecture ofclass I aldolases have shown that the functionalenzyme is about 160 000 Daltons and consists

of four identical subunits.14 37.39 The most thor-oughly investigated class I aldolase, rabbitmuscle aldolase (aldolase A), has been crystal-lised and the three dimensional structuredetermined by x ray diffraction to a resolutionof 1.9 Angstr6ms.40 4' The polypeptide back-bone of the enzyme folds into an alternatingarrangement of a helices and f sheets, whichadopt a basic eight stranded I barrel typestructure, similar to triose phosphateisomerase" 4 and the A domain of pyruvatekinase,44 as well as 14 other enzymes with littlesequence, but significant structural homology(reviewed by Farber and Petsko45). Sincehuman aldolase B has 69% amino acid identityto its paralogue aldolase A, it is likely to assumea similar three dimensional conformation, asdoes rabbit liver aldolase.Using the structural data with the corre-

sponding amino acid sequence,46 the substratebinding residues Lys 146, Arg 148,47 and Lys229 (which forms the Schiff base3) line theactive site pocket which is located in the centreof the f barrel. Even though the spatialconfiguration of the carboxy-terminal tail ofaldolase A could not initially be identified bySygusch et al,40 experiments on the evolutionar-ily conserved terminal tyrosine residue, eitherby carboxypeptidase treatment' 48 or site di-rected mutagenesis of recombinant aldolaseexpressed in Escherichia coli,49 have shown thatit is involved in binding of the substrate,fructose-1,6-bisphosphate. The C terminusperhaps mediates movement of substrates inand out of the barrel structure.40 A more recentmodel based on refined data obtained fromrabbit aldolase A has indicated that the Cterminal residues actually project into theactive site.4'The gene for human aldolase B has been

mapped to chromosome 9q22.3.5° The struc-tural organisation shows nine exons, includingthe first 72 bp exon (nucleotides 924-996),which lacks the initiator methionine codon inexon 2 at nucleotides 5816-5818, to span aDNA segment of 14.5 kb.53 The cloning of thealdolase B cDNA shows that the mRNAencodes a 364 amino acid polypeptide includ-ing the initiator methionine.54 The A and Cisoforms of human aldolase have also beencloned and their cognate genes map tochromosomes 16 and 17, respectively.36 55 Kita-jima et atr6 observed that there are sevencommon conserved and four isozyme groupspecific regions in the sequences.56 The car-boxyl tail shows significant sequence diver-gence, thus the unique functional properties ofthe individual isozymes may be invested in thisregion. Indeed, there is clear evidence tosupport the contention that isozyme function isdependent on changes in this single domain,since sequence analysis of the aldolase locusfrom Drosophila melanogaster shows a singlegene with three alternative 3' terminal exons.57The single primary transcript undergoes alter-native splicing to give rise to the three isozymespecific aldolases with variable carboxy-terminal sequences and differing catalyticproperties." 5

354



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/


Consequences ofhuman aldolase BdeficiencyHereditary fructose intolerance (HFI, OMIM22960) is a recessively transmitted metaboliccondition caused by catalytic deficiency of liveraldolase (aldolase B) (EC 4.1.2.13) in the liver,kidney, and intestine.60The disease, first reported in an adult in

1956 by Chambers and Pratt,6' was typified bythe description of a 24 year old woman who,after taking sugar and fructose complained ofphobic symptoms, faintness, abdominal pain,and nausea; when she took glucose, thesesymptoms were absent, although it was notedthat she did not enjoy the sweet taste. Aftersystematic tasting with a range of sugarsunknown to the patient, it was concluded thatshe had an "idiosyncrasy to fructose". Cham-bers and Pratt suspected, but did not ever for-mally prove, that several of the symptomsnoted in their patient were the result offructoseinduced hypoglycaemia.

Studies of HFI in infants and children soon

followed and showed a more alarming disorderin those affected subjects who are unable toavoid the toxic sugars.62 63 Infants with thiscondition are the most vulnerable to exposureof dietary fructose, especially at weaning.64 65The newborn does not develop any symptomswhile taking breast milk since this contains lac-tose, the disaccharide of glucose and galactose.However, characteristic symptoms ofvomiting,nausea, and sweating, associated with hypogly-caemia and metabolic acidosis, are inducedafter transfer to sweetened milk formulae andsolid foods containing added sugar, as well as

natural fruit and vegetables. If large quantitiesof sugar are consumed, the acute reaction ismore severe and the infant becomes lethargicand may develop seizures or coma.

It has been observed that persistent intake ofthe harmful sugars in childhood leads to a syn-drome of chronic toxicity.66 At this stage alsoirreversible damage to the liver and kidney mayoccur.67 This course eventually leads to cirrho-sis ofthe liver and sometimes death. Clearly themother, or others responsible for infant care,play a critical part in the nutrition of thoseaffected and are usually best placed to protectthe defenceless patient by the early withdrawalof foods and drinks that cause the disease, thusidentifying the harmful items for the person toavoid as they become more independent. Insome countries, carers other than the con-cerned parents may interfere with the nutritionof ailing infants with HFI: the authors areaware of several instances where grandparentshave administered large quantities of honey toinfants under investigation for the diseaseunbeknown to nursing staff and parents inhospital, with catastrophic results.

If the undiagnosed infant survives thedifficult initial period of weaning, the childusually develops a self-protective aversion tofoods which cause distress.6' Voluntary dietaryexclusion, which is refined by trial and errorover a life time, includes restriction of mostsweet tasting foods. The older infant affectedby HFI often proves difficult to feed and recur-

rent ill health and growth failure arecommon.64-66

Adults with the condition, who have survivedas a result of a self-imposed low fructose diet,can remain undiagnosed for many years untilthey come forward in response to articles onthe subject in the public domain or as a resultof their dentist's observations. Several dentalresearchers have noted that children and adultswith HFI have reduced dental caries as a resultof their modified dietary habits.697'A particular hazard for people with undiag-

nosed HFI has been the indiscriminate use offructose infusions as a source of parenteralfeeding72 73; more than 20 fatal or near fatalinstances resulting from this cause have beenand continue to be reported in HFI174-77 (seealso below). These cases have lately beenalmost exclusively restricted to Germany wherethe use of parenteral fructose and sorbitolsolutions has persisted. Although the use offructose as opposed to glucose containingfluids as a source of parenteral nutrition wasoriginally thought to be beneficial for patientswith diabetes mellitus,"8 their use is not onlyfatal in patients with HFI,79 80 but is potentiallytoxic in normal subjects.8" 82 Hence, in mostcountries their availability has been restricted:glucose and lipid preparations are now the pre-ferred sources of energy. In a recent example,medical review and molecular analysis of aldo-lase B genes in surviving relatives of the deadpatient (a German citizen) provided evidencethat led to successful litigation by the wife andfamily of the proband; the hospital authoritiessettled compensation (see data and pedigree infig 2).Once the diagnosis of fructose intolerance

has been established, a strict exclusion dietshould be introduced with the assistance of anexperienced dietician; provided that tissuedamage has not been extensive, normal healthand development returns rapidly.83 84 Careshould be taken to advise on the suitability ofall medicines: sucrose and sorbitol are favouredexcipients and coatings for tablets and are fre-quent components of syrups and suspensionsused to deliver drugs in a palatable form toinfants and children. There are numerousexamples of their harmful effects in patientswith HFI, who should be advised to takesupplements of water soluble vitamins, includ-ing folic acid and vitamin C; "Ketovite" is asuitable preparation for this use. We recom-mend that patients should use warning Medic-Alert bracelets advising on prohibited sugars(especially for the traveller) and on the appro-priate treatment for hypoglycaemia in strickenpatients (glucose or milk for parenteral or oraluse as needed). At the time of writing, theoriginal patient reported by Chambers andPratt6' is alive and well known to the authors.She is in good physical health at the age of 67years and is a grandparent.

CLINICAL INVESTIGATIONS AND PATHOGENESISThe specific pathway of fructose metabolism,which was brilliantly elucidated by Hers,8" isdepicted in fig 1. To summarise, once exo-genous fructose is incorporated into the cell, it

355



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/

356

I_

I,u

Figure 2 Retrospective diagnosis ofHFI by molecular analysis of aldolase B genes. Theunexpected death in a German hospital of a local 42 year old patient (I. 2) with a historyofpeculiar eating habits (disliked sweet foods), who received intravenous infusions offructose and its derivative, sorbitol, to correct dehydration after a minor diarrhoeal illness, aswell as the death of his younger brother (1. 4) under similar circumstances, promptedlawyers representing the widow andfamily to request analysis of aldolase B genes insurvivingfirst degree relatives. Agarose gel ofPCR amplified DNA containing exon 5fragments of the aldolase B gene after digestion with the restriction endonuclease AhaII."Partial cleavage of the 322 base pairfragment into 186 and 136 base pairs confirms thepresence of a single copy of the aldolase B mutant allele A 149P Both daughters, as well asthe surviving brothers, of the dead man (1. 2) are indeed carriers ofHFI suggesting thatthe dead brothers hadfructose intolerance having inherited a single copy of the disease allelefrom each of their consanguineous parents. Dark shading, genotype determinedfrom the gel;light shading, genotype determined by inference; . 1 and L 2, natural death; 1. 2 and 4,premature death.

is rapidly phosphorylated into fructose1-phosphate"6 by the action of fructokinase.88In this specialised pathway, fructose1-phosphate is cleaved by aldolase B, thussplitting the asymmetrical 6-carbon sugar intoD-glyceraldehyde and dihydroxyacetonephosphate.)2 These intermediary triose prod-ucts can be incorporated directly either into theglycolytic pathway (with the eventual forma-tion of lactate and pyruvate, which act as sub-strates for the Krebs' cycle) or the gluconeo-genic pathway to provide a source of glucose, orthe synthesis of glycogen.

Investigations have been conducted to deter-mine which aspects of fructose metabolism are

impaired in patients with HFI. Froesch et alr9incubated liver biopsy specimens with radiola-belled fructose and found that in patients withHFI the incorporation of fructose into glyco-gen was less than 6% of normal. This suggeststhat formation of glycogen from fructose isprincipally catalysed by aldolase B and that theactivity of any alternative pathways for thedirect conversion of fructose 1-phosphate tofructose-1i,6-bisphosphate in human liver tis-sue is limited.

Analysis of blood and urine metabolites inpatients with HFI have shown that when fruc-tose is administered, blood glucose, phosphate,and potassium concentrations decrease; this isaccompanied by increased concentrations ofblood magnesium, urate, alanine, andlactate.63 9( 91 If the concentration of fructoseexceeds 2 mmol/l in the blood, fructosuria, asshown by the appearance of reducing sugars inthe urine, also occurs. The biochemicalchanges observed in HFI appear to be second-ary to the fructose 1-phosphate aldolasedeficiency, affecting the otherwise normalmetabolic pathways of fructose metabolism inthe liver and accessory tissues. Although thefructosuria of HFI patients exceeds that innormal subjects after sugar loading, only afraction of the fructose ingested is excreted bythis route. This indicates that fructose ismetabolised at extrahepatic sites92: the sugges-tion that hexokinase in leucocytes, erythro-cytes, and adipose tissue may contribute to thedisposal of excess plasma fructose has beensupported by metabolic labelling studies usingsamples of rat epididymal fat.93 The biochemi-cal changes described here immediately ac-company the acute symptoms of sugar toxicityin HFI.6 65When aldolase B activity is deficient, the

rapid phosphorylation of fructose creates abuild up of fructose 1-phosphate with depletedintracellular inorganic phosphate and ATP inthe organs where the specialised pathway offructose metabolism is operative.'( 94 Thereduced Pi concentration activates adenosinedeaminase and the xanthine oxidase pathways,with degradation of purine nucleotides to uricacid.95 97 Reduced intracellular concentrationsof inorganic phosphate lead directly to theallosteric activation of adenosine deaminase.9"The increased formation of uric acid leads tohyperuricaemia, a metabolic effect of fructoseadministration observed to a limited extent innormal subjects.9 More recently, it has beenobserved that the hyperuricaemic effect is morepronounced in HFI heterozygotes than in con-trol subjects,99 suggesting that carriers of thisdisorder may be predisposed to hyperuricae-mia and gout. Reduced intracellular concentra-tions of ATP appear to account for the releaseof Mg2+ ions"'0 by dissolution of the Mg-ATPcomplex, so causing the hypermagnesaemiaobserved in HFI patients following challengewith fructose.63 101 High concentrations ofunmetabolised fructose 1-phosphate inhibitfructokinase action by negative feedback, thuspreventing further incorporation of fructosewhen metabolism by the aldolase B pathway issaturated.'02 As a result, transient fructosaemia(an outmoded term for hereditary fructoseintolerance) and fructosuria are observed.'03 ")9The hypoglycaemic response to fructose

experienced by patients with HFI is attributableto the impaired hepatic breakdown of fructose1-phosphate. Experiments carried out in pa-tients with HFI during fructose inducedhypoglycaemia clearly show reduced interpran-dial plasma glucose release'04 without evidenceof hyperinsulinism. The failure of parenteralglucagon to correct the hypoglycaemia indicates

Ali, Rellos, Cox

.i :-.;, -



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/


Figure 3 Aldolase

anaemia. Agarose gDNA obtainedfron2) and after the tra

subject (lane 8), a )homozygous forAextractedfrom a saiafter the transplant.heterozygous for theleucocyte cells with

recipient had bone

chimaerism, combinthe transplant, confirecovery of the apla(courtesy of Drs PDebri, Paris).

defective glycogenolysis. Although infusionsof galactose raise the blood glucoseconcentration (indicating that the phospho-glucomutase and galactokinase pathways areintact),9' "' a lack of response to dihydroxyac-etone shows that gluconeogenesis is alsoimpaired. The exact mode of inhibition of gly-cogenolysis is unknown: in vitro studies suggestthat increased concentrations of fructose1-phosphate, combined with the decrease in

ship is lacking. The possible involvement offructose 1-phosphate in the regulation andsynthesis of glycoproteins has received littleattention but the high concentrations of thisester that occur in HFI tissues after fructoseexposure might also have hitherto unsuspectedeffects on hepatic glycan metabolism, includ-ing the synthesis of desialylated plasma trans-ferrin isoforms." 6

inorganic phosphate, block glycogen break- GENETIC STUDIESdown at the level of phosphorylase.'06-'08 Fruc- Information obtained from the cloning andtose 1-phosphate also prevents the formation of characterisation of the wild type humanthe gluconeogenic intermediates, fructose-1,6- aldolase B gene 54 allowed Cross and col-bisphosphate and glucose 6-phosphate, by leagues to clone the chromosomal aldolase Bcompetitive effects on aldolase A and glucose gene from a well characterised patient suffering6-phosphate isomerase, respectively.'09 110 In from hereditary fructose intolerance, who hadthe absence of a fructose challenge, no such an intact but functionally inactive enzymemetabolic inhibition would be expected and aldolase B polypeptide expressed in liver andthis is certainly supported by the observation intestinal mucosa."7 118 Sequence analysis ofthat patients with HFI tolerate prolonged fast- the entire mutated gene showed a single baseing. substitution in exon 5 of aldolase B from theImpairment of gluconeogenesis after fruc- patient. Molecular analysis of genomic DNA

tose intake in patients with HFI combined with using the polymerase chain reaction followedfructose 1-phosphate induced activation of by restriction enzyme digestion showed thatpyruvate kinase"' results in accumulation of the patient was homozygous for this mutantthe Krebs' cycle precursors, alanine, lactate, allele. This mutation, a G-*C transversion inand pyruvate. This contributes to amino the first base of codon 149, is inferred toacidaemia, as well as metabolic acidosis.82 112 replace the normal alanine by a proline residue;Impaired function of the proximal renal tubule the variant allele was hence designated Al 49P.aggravates the acidosis and results in general- The homozygous genotype, A149P, was foundised amino aciduria, phosphaturia, and bicar- only in association with symptomatic patientsbonate wasting."3 1"4 As a result, adolescents with HFI, who had inherited one copy of thepresent with metabolic bone disease and stunt- allele from each parent, and confirmed that theing of growth.66 Defective acidification of the condition segregated in the families as ex-urine, as well as phosphaturia and amino pected for a recessive disease.68 1" It alsoaciduria, represents an acquired form of the provided strong evidence that defects in theFanconi syndrome that has been well docu- aldolase B gene cause HFI. Facile detection ofmented in HFI."' The cause of the abdominal the A149P mutation, which creates a novelpain'15 that is a feature in HFI remains recognition site for the restriction endonucle-unexplained. This pain, which follows inges- ase AhaII and its isoschizomers, has been usedtion of the offending sugars within a few for molecular diagnosis68 and confirmatoryminutes, may reflect autonomic afferent activ- testing after large scale genotyping by alleleity induced by release of purine nucleotides, specific hybridisation."'9 Indeed we have usedloss of energy charge, or increased concentra- this procedure to ascertain the success oftions of gluconeogenic precursors such as engraftment following the use of allogeneiclactate, but proof of a cause and effect relation- bone marrow transplantation from a donor

with established HFI into his HLA identical2 3 4 5 6 7 3 9 1 but A149P heterozygous brother who suffered

from acquired aplastic anaemia (fig 3).Since this original investigation, molecular

analysis of human aldolase B genes using thepolymerase chain reaction and DNA sequenc-ing has to date identified 22 genetic lesions inpatients with HFI (table 1). The detection ofeach mutation in genomic DNA using PCRbased methods has allowed screening to be

B genotyping after bone marrow transplantation for severe aplastc carried out in patients with uncharacterised-el ofAhaII restriction digests using amplified exons S of aldolase Bfrom mutant alleles of aldolase B to confirm a puta-n peripheral blood of the HFI donor (lane 1), the recipient before (lane tive diagnosis ofHFI based on clinical groundsrnsplant at 1, 3, 18, and 21 months (lanes 3, 4, 5, and 6), a healthy alone. Molecular diagnosis by analysis of theheterozygote carrier ofA149P (lane 9), and an HFI patient'49P (lane 10). Lane 7 represents AhaII digestion of amplified DNA aldolase B gene has the obvous advantage thatline mouthwash sample obtainedfrom the transplant recipient 21 months it avoids invasive tests including theBefore transplantation, the donor was homozygous and the recipient intravenous fructose tolerance test and liver oraldolase B gene mutation A149PR The proportion of the recipient's

the A149P allele increased after the transplant, illustrating that the itestial biopsy followed by enzymatic assay.marrow derived cells from the donor. This observation of blood Genetic studies have also provided informationzed with the absence of any recurrent disease symptoms fouryears after about the distribution of mutant aldolase Birms that the engraftment has been curative and that spontaneous alleles in different populations.tstic state had not occurred as a result of cytotoxic conditioning therapy A establishS Rohrlich and E Vilmer, Service Hemato-immunologie, Hopital Robert As established by Cross et al'20 and Cox,'

the most prominent allele causing HFI in

357

322

186



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/

358

Table 1 Human aldolase B gene mutations in hereditary fructose intolerance

Trivial name* Systematic designationt Mutated codont Ethnic distribution Ref

C134R c.343T-*C C135R Amerindian (private) 135W147R c.441T-4C W148R Angloamerican (private) 133A149P c.448G-*C A150P European (widespread) 117A174D c.524C-*A A175D S European (widespread) 120L256P c.770T-4C L257P Italian (private) 149R303W c.910C-+T R304W Turkish/Italian (rare) 130, 132N334K c.1005CG-G N335K E European (widespread) 127A337V c.1013CG-T A338V Turkish/Finnish/Swiss (rare) 130M-1T c.2T-GC MIT Italian (private) 136R3ter c. lOCG-T R4X Turkish (private) 134R59ter c.178CG-T R60X Austria (rare) 134, 135Y203ter c.672T-4A Y204X Italian (private) 136C239ter c.720GC-A C239X Japanese (private) 138G-GC,5' intron 5 IVS 5+1G-sC New Zealand (private) 130, 123G-4A,3' intron 6 IVS 6-1G-sA French (private) 134Q20AA c.62 del A Q21 del Italian (private) 132L288AC c.865 del C L289 del Sicilian (private) 120G12/A3 IVS2-1G del GGTA G38T39 del Swiss (private) 137A4/A4 c.360-363 del CAAA N120 K121 del European (widespread rare) 128, 133A7+1 3' intron 8 IVS8-1G del GGCTAAC ins G A334 N335 del American (private) 164G1O/A 6,7 g.9912-10836 del N181-G267 del Swiss (private) 137F13/A 4,5 g.7516-9165 del L109-5160 del French (private) 137

*Trivial name represents the mutated sequence ofthe inferred translation product after cleavage ofthe initiator methionine residue; similarly the codon mutation refersto the initiator methionine numbered as position (+) 1.tMutations have been assigned according to current guidelines (Antonarakis SE (1996,1997) website ariel.ucs.unimelb.edu.au:80/-cotton/antonara.htm; Hum Mutat(in press)).

patients of European descent, A149P (ac-counting for 65% of European HFI alleles sofar studied), has a wide distribution amongEuropean populations, including AshkenaziJews. It has been found to be prevalent inNorth America, where it accounts for 55% ofHFI alleles studied'22 and has recently beenfound in New Zealand.'23 This allele is the mostfrequent cause of the disease in populations ofnorthern European descent, accounting forover 85% of mutant aldolase B alleles that havebeen studied in the UK; it therefore has a pow-erful discriminatory capacity in relation to HFIdiagnosis.68 120 These findings provide evidencethat the mutation arose early during the evolu-tion of modern human populations and possi-bly before the European expansion thatcharacterised the late Bronze Age.'24 TheA174D allele, which is also widespread,'20accounts for 14% and 1 % of HFI alleles so farstudied in Europe'2' and North America,'22respectively. It is particularly frequent incentral and southern European locations.'2' 125A study to investigate haplotype linkage

analysis of A149P alleles, using allele specificoligonucleotides, has shown absolute associ-ation with informative linked aldolase Bpolymorphisms in intron 8 (c->t at nucleotide84 and a-*g at nucleotide 105) and providesconvincing evidence that the mutation arose ona single ancestral chromosomes and spreadthroughout populations by a combination ofmass migration and random genetic drift.'26These two biallelic single base pair polymor-phisms were found to be in absolute linkagedisequilibrium between themselves and areinformative, since 47% of control subjects werefound to be heterozygous at these loci. Allelespecific hybridisation in 15 HFI patientshomozygous for the A149P aldolase B muta-tion showed that they were also homozygousfor the intron 8 84T/105G alleles. These find-ings accord with the prevalence of these allelesin defined populations'2' allowing speculationas to their origin. The data on haplotypelinkage thus indicate that parents who appear

not to be consanguineous, but give rise to off-spring with the A149P homozygous aldolase Bgenotype, are thus likely to be related by blood,albeit distantly.Another mutant allele of human aldolase B

associated with HFI, N334K,'27 occurs princi-pally in central and eastern Europe, with possi-ble origins in Balkan populations.'2' The intra-genic deletion mutation in exon 4 of thealdolase B gene, A4, which was originallydescribed in a British patient with HFI,128 129has since been identified in a German patient'30as well as six unrelated families with HFI fromItaly'3' 132 and a unique extended NorthAmerican pedigree of Swiss-Germanancestry.'33 Hence, the A4 deletion appears tobe widespread (though not as frequent asA149P) among European populations. Thealdolase B mutation A337V has been shown tooccur in three independent isolates.'30 Origi-nally, the allele was found in homozygous formin a patient from Turkey, who was the offspringof a consanguineous marriage; after screeningfor this mutation, it has been detected in acompound heterozygous form in familiesaffected by HFI from Switzerland (fig 4) andFinland. More recently, the aldolase B mis-sense mutation R303W has been identifiedindependently in a patient from Italy andTurkey.130 132

It is notable that the nonsense mutation inthe aldolase B gene, R59ter, found in anAustrian patient with HFI134 has also beenreported in an Italian-American and a NativeAmerican from British Columbia, who alsohave the disease.'35 Although isolates of A4,A337V, R303W, and R59ter may each bederived from a single ancestral event and havespread by genetic drift, similar to the mutantalleles A149P and A174D, they may havearisen recurrently and more recently as a resultof mutation at "hot spots" in the aldolase Blocus. Indeed the mutations giving rise toA337V, R303W, and R59ter represent a C-Tbase transition (G-4A in the antisense strand),the most frequently observed point mutation in

Ali, Rellos, Cox



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/


A

369 -

246 -

123

B

LIpD St( 1 2 3 4 5

369

246--8

123~~~~~~~~~~~1Figure 4 Compound heterozygosity for aldolase Bmutations A337VandA149P in afamilyfromSwitzerland with HFI. (A) Agarose gel ofMscI digests ofPCR amplified exons 9from a healthy subject (lane 1) andthe Swiss family members (lanes 2-5). Cleavage of the 205base pairfragment into two fragments of 111 and 94 basepairs indicates the presence of the A337V mutation. (B)Agarose gel ofAhaII digests ofexon S fragments of aldolaseBfrom an A149P homozygote HFI patient (lane 1) andthe Swiss family members (lanes 2-5). Digestion of the 322base pair product into 186 and 136 base pairs signifies thepresence of the A149P allele. The mendelian segregation ofthe A337V allele in the family and its association with thedisease phenotype only in subjects with an additionalaldolase B mutation on the other allele, together with theobservation that no other mutation has been identified incis, indicates that Ala 337-* Val inactivates the aldolase Bmolecule.

eukaryotic DNA. Similarly, the event whichgives rise to the A4 allele may represent a hotspot for recurrent frameshift mutations. TheCAAA deletion occurs in a run of repeatedsequence, suggesting that it may have beenmediated by recurrent DNA polymerase slip-page resulting in mispairing of strands duringreplication or occurred at a single ancestralevent. However, in the absence of sufficientnumbers for haplotype analysis, this questioncannot be answered with certainty.

All the other genetic lesions in human aldo-lase B which have so far been identified andgive rise to the disease (table 1) have not beenconvincingly detected in more than one coun-try and so may represent unique mutations thatremain confined to the local community ("pri-vate" mutations). Although CpG dinucleotidesmay represent mutational "hot spots" in genes,of the 15 base substitutions in the human aldo-lase B gene mutated in HFI, only three areC--T transitions and one is caused by a G-Atransition.From all the data that have been published

on the identification of aldolase B mutations inpatients with HFI, it has generally beenobserved that patients from non-consanguineous pedigrees tend to be ho-mozygous for the prevalent aldolase B muta-

tions, A149P, A174D, or N334K, or have acompound aldolase B genotype, having inher-ited at least one of these alleles.'2' The rareexceptions to this observation have been afamily from Italy, where symptomatic subjectshad the aldolase B genotype M-1T/Y203ter,131a pedigree of Italian-American descent wherethe affected patient had the R59ter/C 1 34Rgenotype,'" and a single isolate from a NorthAmerican kindred of Swiss-German ancestry,where the patient had a compound genotype,A4/W147R.'` Where consanguineous mar-riages are frequent they contribute towardshomozygosity for rare mutant alleles of aldo-lase B in HFI, as shown in several Turkishpedigrees. 30 114 There are other illustrativeexamples: in a small community from Sicily theframeshift mutation L288AC was identified,1'20the mutation R303W was present on both alle-les of an Italian patient born to consanguineousparents,'32 the aldolase B lesion, A3, which cre-ates a 4 bp deletion at the junction betweenintron 2 and exon 3, so removing the conservedAG splice site recognition signal, was found inthe homozygous state in a member of a largeconsanguineous family affected by HFI fromSwitzerland,'37 and a female Japanese childwith HFI, whose parents were consanguineous,was homozygous for the mutation C239ter.'39

Several null alleles of aldolase B have beenidentified. These include a translation initia-tion mutation, nonsense mutations, splice sitevariants, and intragenic deletions of 1 bp to1.65 kb (table 1). Since a significant proportionof the mutant alleles of aldolase B associatedwith HFI represent mutations causing synthe-sis of missense enzyme variants,"8 129 139 140)identification of rare null alleles present inhomozygous form allow the human aldolase B"knockout" to be compared with the pheno-type where residual hepatic fructose1-phosphate aldolase activity persists. Patientswho have inherited two null alleles of aldolaseB'20 134 136 138 appear to be healthy, provided theyavoid fructose, and are able to withstandstarvation. This indicates that liver aldolase Bactivity is not critical for gluconeogenesis in theinterprandial state. Aldolase B is thus acomponent ofan accessory pathway that allowsefficient metabolic assimilation of specificdietary sugars as part of the adaptive responseto rapid nutritional changes; gluconeogenesisand glycolysis principally depend upon theubiquitous aldolase A.As far as we can establish, the presence of

homozygosity for null alleles or missensevariants does not determine the clinical severityof HFI. The symptoms and extent of organdamage appear principally to depend on thesubject's immediate nutritional environment.This is in part reflected by their level of educa-tion as well as the dietary habits of their cultureand the alacrity with which their parentsrecognised feeding difficulties presenting dur-ing infancy and childhood. Whatever thereason, in nearly all instances, the aldolase Bmutations cause profound deficiency of fruc-tose 1-phosphate aldolase activity and areeffectively null mutations. In only one instancehas asymptomatic HFI been associated with

359



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/

Ali, Rellos, Cox

disease solely under conditions of extremefructose loading.'4' Unfortunately, neither de-tailed enzymology nor molecular analysis oftheputative mutant aldolase is available from thispatient.

MOLECULAR STUDIESSince human aldolase B is expressed only in therelatively inaccessible fructose metabolisingtissues (the liver, kidney, and intestine), it is noteasy to study precisely the effects of the muta-tions on enzymatic activity in situ. Liver biopsyspecimens, although useful for the diagnosis ofHFI by assay of fructaldolase activity,60 cannotbe obtained without appreciable risk to thepatient and usually provide too little tissue fordetailed characterisation of mutant enzymes.

Nevertheless, studies using antibodies spe-cific to each isozyme of human class I aldolaseto investigate histological sections and foranalysis of tissue extracts by immunoblottingshowed that some patients with HFI expresspolypeptides that bind to aldolase B specificantibody."'8 129 143 Experiments using doubleimmunodiffusion gels showed that antiserumagainst human aldolase B partly activated thealtered enzyme in several liver extracts frompatients with HFI."0 An independent study byCox et al,"8 using a radioimmunotitration pro-cedure showed that the aldolase B frompatients had a reduced affinity for the antibody,suggesting that the defect created a structuralvariant of aldolase B in these patients, withimpaired catalytic function. Furthermore,these authors succeeded in purifying mutantaldolase B from tissue extracts obtained from apatient with HFI using immunoaffinity meth-ods. They showed that the purified enzyme wascatalytically inactive and possessed an identicalelectrophoretic charge to the wild type enzyme,although it had an apparent subunit molecularsize of 39 100 daltons (wild type is 38 000daltons). 118 Subsequently, molecular analysis ofaldolase B genes showed that the mutationcausing HFI in this pedigree encoded theenzyme variant, Ala'49-Pro (Al49P)."7With the introduction of gene cloning to iso-

late human aldolase B cDNA and model host/vector expression systems, catalytically activerecombinant human aldolase B has beensuccessfully expressed in E coli.'2' 114 Clearly,the missense variants of human aldolase B,obtained by site directed mutagenesis, that areresponsible for disease are the most interestingto study at the molecular level using thesetechniques because they shed light on howsubtle changes in critical residues of the proteinaffect its function. Hitherto, eight aldolase Bmissense mutations have been identified inpatients with HFI (table 1). All mutatedresidues are invariant in the class I aldolases,except the alanine at position 149 which is Bisozyme specific: aldolases A and C have acysteine at this position.' ' This suggests thatthe regions that harbour these mutations playan important role in the catalytic function orstructural integrity of the normal tetramericenzyme. Using in vitro mutagenesis withsynthetic DNA oligonucleotides, it is possibleto create mutant plasmids which encode the

single amino acid substitutions so that theproperties of the normal and mutant forms ofrecombinant human aldolase B can be sub-jected to detailed examination.So far the disease causing aldolase B variants

C134R and A149P have been synthesised in Ecoli, to understand further the effects of thenatural mutations on the enzyme molecule.Brooks and Tolan,'35 who expressed the mutantaldolase B corresponding to the single basesubstitution C 134R in E coli, showed thatalthough the mutant enzyme retained residualcatalytic activity towards the fructose1-phosphate substrate, the reduced levels ofexpression compared with wild type in thishost/vector system suggested that structuralinstability of the partially active enzyme variantmay be the cause of the disease phenotype. TheA149P mutation appears to have a drasticeffect on enzymatic activity and structure ofaldolase B as indicated by immunologicalstudies in tissues obtained from homozygouspatients."7 118 The expressed protein is inactivetowards its specific substrate, fructose1-phosphate, with residual activity towardsthe fructose- 1,6-bisphosphate substrate."'2The schematic diagram of the three dimen-

sional structure of human aldolase A (which ishighly homologous to aldolase B) can be usedto model the missense mutations in aldolase Bassociated with HFI (fig 5). The mutationsC134R, W147R, A149P, and A174D on exon5 all appear to be located in the vicinity of thesubstrate binding pocket and so are likely todisrupt important residues, including the C-1phosphate binding residues, Lys 146 and Arg148, in the active site of the enzyme.40 145 46 Inaldolase A, x ray diffraction of a resolution of1.9 A indicates that Lys 146 is sufficiently closeto function as a proton donor or acceptor dur-ing carbinolamine intermediate formation,41 afunction inhibited by engineered mutations ofthis residue.'46The residues implicated in the N334K and

A337V mutations (exon 9) occur in an a helixand affect the mobility of the conformationallyflexible carboxy-terminal tail, a region which isthought to cover and mediate access of thesubstrate into the active site pocket.40 Thisregion has long been considered critical forfunctional differentiation of the aldolase iso-zymes A, B, and C.56 147 148 It has been claimedthat the wild type residue in R303 (exon 8)participates in binding of the phosphate at the6-carbon position of fructose- 1,6-bisphos-phate, although this supposition is based onmodelling of the aldolase A molecule ratherthan definitive x ray structural determinationsand analysis.'45 The change in charge and bulkof the side chain associated with this mutationis likely to affect the structural integrity of themolecule. An entirely different picture of therole of R303, however, has emerged from thesolution of the aldolase A structure at the highresolution of 1.9 A.4' It interacts with theC-terminal region residue Glu 354 and forms ahydrogen bond with the C, carbonyl oxygen ofbound dihydroxyacetone phosphate which isderived from the C, moiety of the hexose

360



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/


Figure 5 Schematic diagram of three dimensional structure ofhuman aldolase A at 2.04ngstrdm resolution (after Gamblin et al, 45 with permission). The a helices and ,B sheetsare represented as barrels and arrows, respectively. The carboxy-terminus, which is depictedas bold, is postulated to extendfrom the tail of helix H2 to the centre of the a/fl barrel. Theposition of amino acid substitutions identified in the human aldolase B gene ofpatients withHFI are mapped onto the diagram.

substrate and stabilises the keto form of theincoming substrate at the active site.Tetramer formation required for functional

aldolase B involves close contacts betweenhydrophobic residues40; it may be the case thatthe exon 7 mutation, L256P,'4' which appearsto lie on the outer surface of the molecule dis-rupts this hydrophobic interaction, possiblyresulting in a failure oftetrameric assembly andformation of aldolase subunits which, althoughcatalytically active, are unstable.'50 151 However,the exact consequences of all these mutationson the catalytic function of the aldolase B iso-zyme and integrity of the molecule will requiresystematic analysis of this particular humanenzyme using protein engineering techniques.Detailed investigations of the catalytic functionand definitive structure of the engineered aldo-lases, informed by computer modelling studies,will doubtless be critical for a full mechanisticunderstanding of this ancient enzyme family.

DiagnosisAlthough HFI may be suspected in people whoshow aversions to sugary foods, confirmationof the diagnosis is important, particularly forthe infant, so that a strict exclusion diet can beprescribed and the tissue injury and growthretardation can be avoided.65 83 84 Confirmatorystudies hitherto have relied either on directassay of fructaldolase activity in tissue biopsyspecimens" 142 152 or inducing the characteristicbiochemical changes after carefully controlledadministration of fructose (0.2-0.25 g/kg bodyweight) by intravenous infusion.89 153 Morerecently, 31P nuclear magnetic resonance spec-

troscopy has been used successfully to show anincrease in sugar phosphates and decrease ininorganic phosphate in the liver ofHFI patientsafter a fructose load.99 Indeed this method, andenzymatic assay of intestinal biopsyspecimens,'52 have been the only procedures

known to date that allow the detection of other-wise asymptomatic carriers of HFI. Never-theless, these techniques are either cumber-some, expensive, or have obvious risks; this,combined with the pain and inconvenience thesubject may have to suffer, fully justifies effortsto develop direct methods for genetic diagnosisof HFI.The discovery of the polymerase chain

reaction'54 has allowed direct DNA analysis ofaldolase B gene sequences, obtained fromsomatic cells of patients, to be developed for anon-invasive diagnosis of HFI. Although threemutations in the human aldolase B gene,A149P, A174D, and N334K, and more re-cently A4, A337V, R303W, and R59ter havebeen found to be widely distributed, a morecomplete picture with respect to the global dis-tribution of aldolase B mutations is required,otherwise genetic screening for HFI will belimited to specific populations in which therepresentative mutant alleles can be predictedwith confidence. The pattern of mutationfrequencies appears to conform to a prototypethat has been found with several other genes: afew are prevalent but most are rare. The preva-lent mutations, for example, A149P, A174D,and N334K, differ in frequency betweendifferent populations as with phenylketonuria.The high prevalence of A149P in the UK inparticular is very remarkable and probablyexplained by a founder effect early in the colo-nisation of these islands. For diagnostic pur-poses in other populations, rare mutant allelesof aldolase B can be overlooked leading to mis-diagnosis.To detect whether a suspected person has a

molecular lesion in the aldolase B gene, a sim-ple protocol has been adopted in ourlaboratory.'30 Initial screening for the previ-ously identified aldolase B mutations A149P,1"7A174D (Ala 174->Asp), and N334K,'20 127which are a prevalent cause of HFI, isconducted on genomic DNA, using the PCRcombined with restriction enzyme digestion oroligonucleotide hybridisation. A single mutantallele in the symptomatic patient confirms thediagnosis. Subjects who have wild type allelesat these loci and if their parents were unrelatedby blood were studied further only if a positiveintravenous fructose intolerance or enzymatictest had been obtained previously. However,patients from consanguineous marriages whocomplained of typical symptoms were investi-gated to find the mutant allele common to theirparents. Direct sequencing of aldolase B exons,including the splicing signals, was used forscanning the gene to detect any unidentifiedmutations because it is systematic; clearly, asautomated sequencing becomes generallyavailable, it will offer an increasingly attractivemeans of effecting an accurate diagnosis of thisdisease. Once both mutant alleles of aldolase Bwere identified in a proband, this informationcould be used in the neonatal screening of sibsfor HFI using either cord blood or Guthriecard material from the child as a source ofgenomic DNA template and before exposureto fructose. By these means we have been able

361



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/

Ali, Rellos, Cox

to counsel parents before weaning their off-spring onto potentially harmful foods.'

Frequency ofHFI in the populationAn independent study from the UniversityChildren's Hospital, Zurich,'55 to estimate theprevalence of hereditary fructose intolerance,recorded five patients with HFI out of nearly100 000 live births in a single centre over a fiveyear period. Thus, the frequency estimate forHFI was approximately 1 in 20 000; however,the confidence limits for such an estimatebased on small numbers of affected subjectswould be wide (approximately 1 in 11 000 to 1in 100 000). Even so, paediatricians and physi-cians consider the condition to be very rare,despite growing evidence that mutant alleles ofaldolase B causing the disease may be morefrequent than at first realised.

(1) As a rule, the parents of affected patientsare not consanguineous. For rare autosomalrecessive diseases, clearly, as noted byGarrod,156 consanguinity would be expected.

(2) There are numerous reports of parent tooffspring transmission resulting from the mar-riage between affected homozygotes withasymptomatic heterozygous carriers. "' 152157-159

Because this disorder is inherited as a recessivetrait, this observation signifies an appreciablefrequency ofmutant alleles in the population atlarge. Indeed, several large pedigrees affected byHFI have been reported that suggest, in theabsence of consanguinity, that mutant alleles ofaldolase B are prevalent.9' 152 160 In onefamily,'33 10 affected subjects were identifiedbecause of the independent segregation of fourmutant alleles (A4, W147R, A149P, A174D);there was an example ofpseudodominant verti-cal transmission of disease and all affectedpatients were compound heterozygotes; noconsanguinity was identified.

(3) After surviving the initial traumaticperiod of weaning and infancy, many affectedsubjects as adults adjust their eating habitsaccordingly and can live a normal life. In thisway, they escape formal diagnosis. However,people continue to come forward in responseto articles in the lay press about thedisorder.68 123

(4) The effects of administration of fructosebased solutions during surgery have resulted inat least 17 deaths of patients not previouslyknown to be suffering from HFI.7s 76 133 136 161163

Since heterozygotes, who are asymptomatic,manifest no obvious advantage compared withnormal subjects, there is no known mechanismto account for the prevalence of aldolase Bmutations, apart from the observation thataffected homozygotes have a reduced incidenceof dental caries. This may have favourably con-tributed to fitness in earlier times. It has beensuggested that although mutant alleles of aldo-lase B may have been accumulating over theyears as a result of genetic drift, homozygoteswould have escaped detection when sugar,honey, and fruit were scarce. Only the increas-ing consumption of sugars in the western dietover the last century has led to the emergenceof HFI as a prevalent disorder.'2'

Given the potentially severe consequences ofuntreated fructose intolerance and the value ofdiagnosis as well as the efficacy of strict dietarytreatment, there is a case for determiningwhether mass neonatal screening is justified forthis avoidable nutritional disorder. In this labo-ratory we have recently conducted a pilotstudy, to obtain an estimate of the populationfrequency of the mutant aldolase B allele,A149P, by systematic analysis of DNA ob-tained from Guthrie blood spots taken atbirth."9 Blood spots obtained by heel prick ofthe newborn are collected routinely on a cardand tested for various preventable diseases,including phenylketonuria (PKU) and con-genital hypothyroidism. The dried blood spotsprovide an unbiased population for screening,which is also relatively inexpensive for themolecular diagnosis of HFI. Our study identi-fied 27 A149P heterozygotes from 2050 unse-lected subjects born within a nine monthperiod between 1994 and 1995 in the EastAnglian region of the UK (frequency (2pq)-1.3%). These data allow for a frequency (q2)of 1 in 23 000 homozygotes to be predicted forthis allele"9 (giving a disease frequency of 1 in18 000 assuming the A149P variant accountsfor -80% of mutant alleles).68 121 Clearly thishas important implications for establishinginterventional programmes for HFI.

It is noteworthy that whereas the mutantaldolase B allele A149P, which is widespread inwhite populations, has been shown to accountfor over 85% of HFI alleles which have so farbeen studied in the UK'12 and so is diagnosti-cally significant, diverse alleles give rise to thedisease in other populations, such as the Italianpopulation. 1215131 136 149 So far, 10 aldolase Bmutations, M-1T, AA20, A4, A149P, A174D,Y203ter, L256P, L288AC, R303W, andN334K, have been identified to be associatedwith HFI in Italy, which would make for acomplex diagnostic strategy based on mutationanalysis. Forthcoming advanced automatedmethods for detecting multiple gene defectswill be ideally suited for mass screening inpopulations such as the United States, Italy,and Switzerland with diverse mutations inaldolase B.120 122The aim ofneonatal diagnosis is to introduce

therapy before disease symptoms manifestthemselves, thus maintaining the health andlife quality of the affected subject. Althoughcarrier detection has no direct health implica-tions for the child, counselling and DNAtesting might be justified when decisions aboutreproduction are made in later life, especially inpopulations where consanguineous marriagesoccur with appreciable frequency.

ConclusionsThe nature of the molecular defect in the aldo-lase B genes of patients with HFI does notappear to affect the phenotype appreciably;however, the individual dietary experienceappears to be correlated with disease severity.The widespread mutant alleles of aldolase Bwhich have so far been identified in certainregions of the world prompt further studies todetermine the frequency of aldolase B mutant

362



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/


alleles in other populations and hence deter-mine the extent to which neonatal screening forhomozygotes by genetic testing is justified. Toimplement a general population screening pro-gramme for HFI, the benefits of such aprogramme must outweigh the cost of settingup the whole procedure and there should beprovision for a considerable educational andcounselling effort. Ethical considerations mustalso be taken into account includingconfidentiality issues, since future access coulderroneously be considered to affect the insur-ance status of the homozygous or compoundheterozygous person who, after all, would havean avoidable clinical disorder.The use of expression vectors in E coli has

enabled recombinant aldolases to be generatedin vitro, thus opening the way for experimentalmanipulation. Important studies are inprogress to answer questions about the precisemechanism by which the human aldolase Bfolds, from its linear primary translation prod-uct to the active tertiary conformation and sta-ble homotetramer. More importantly, physicalstudies will be able to explore the tantalisingrelationship between regional structure and thespecialised biological function of each isozyme.The naturally occurring missense mutationsassociated with HFI are useful in that theyfocus attention on critical residues in regions ofthe molecule which are required for aldolase Bcatalytic function. Detailed biochemical analy-sis of purified mutant aldolases B, includingtheir crystal structure in the presence of the fiveknown substrates, is likely to reveal much aboutthis ancient but highly differentiated modelenzyme system.

Research conducted in the authors' laboratory was supported bythe Medical Research Council and The Wellcome Trust. Wethank Joan Grantham for secretarial assistance and our manycollaborators who have referred diagnostic samples for analysis.We also wish to thank Dr Michael Camilleri and Dr NicholasCross for their earlier contributions to the study of this disease.

1 Rutter WJ, Woodfin BM, Blostein RE. Enzymic homology.Structural and catalytic differentiation of fructose diphos-phate aldolase. Acta Chem Scand 1963;17:5226.

2 Van der Heiden BS, Meinhart JO, Dodson RG, Krebs EG.The preparation and properties of crystalline yeastaldolase. J Biol Chem 1962;237:2095-8.

3 Horecker BL, Pontremoli S, Ricci C, Cheng T. On thenature of the transaldolase-dihydroxyacetone complex.Proc Natl Acad Sci USA 1961;47:1949-55.

4 Blostein R, Rutter WI. Comparative studies of liver andmuscle aldolase. II 44. Immunochemical and chromato-graphic differentiation. J Biol Chem 1963;238:3280-5.

5 Penhoet E, Rajkumar T, Rutter WJ. Multiple forms of fruc-tose diphosphate aldolase in mammalian tissues. Proc NatlAcad Sci USA 1966;56:1275-82.

6 Penhoet EE, Kochman M, Rutter WJ. Isolation of fructosediphosphate aldolases A, B and C. Molecular and catalyticproperties of aldolase C. Biochemistry 1969;8:4391-402.

7 Munnich A, Besmond C, Darquy S, et al. Dietary and hor-monal regulation of aldolase B gene expression. 7 ClinInvest 1985;75:1045-52.

8 Adelman RC, Spolter PD, Weinhouse S. Dietary andhormonal regulation of enzymes of fructose metabolism inrat liver. J Biol Chem 1966;241 :5467-72.

9 Rosensweig NS, Stifel FB, Herman RH, Zakim D. Timeresponse of diet-induced changes in human jejunalglycolytic enzymes. Fed Proc 1969;28:323.

10 Arnold H, Pette D. Binding of aldolase and triosephosphatedehydrogenase to F-actin and modification of catalyticproperties of aldolase. EurJ Biochem 1970;15:360-6.

11 Weiss TL, Zieske JD, Bernstein IA. Reversible microsomalbinding of hepatic aldolase. Biochim Biophys Acta 1981;661:221-9.

12 Sato JI, Tsutsumi KI, Ishikawa M, Ishikawa K. Dietaryregulation of aldolase isozyme expression in rat intestinalmucosa. Arch Biochem Biophysl 987;254:116-23.

13 Agius L. Substrate modulation of aldolase B binding inhepatocytes. Biochem 1996;315:651-8.

14 Lebherz HG, Rutter WJ. Distribution of fructose diphos-phate aldolase variants in biological systems.Biochemistryl969;8: 109-21.

15 Hommes FA, Draisma MI. The development of L- andM-type aldolases in rat liver. Biochim Biophys Acta1970;222:251-2.

16 Numazaki M, Tsutsumi K, Tsutsumi R, Ishikawa K.Expression of aldolase isozyme mRNAs in fetal rat liver.EurJ Biochem 1984;142: 165-70.

17 Schapira F, Dreyfus J-C, Schapira G. Anomaly of aldolase inprimary liver cancer. Nature 1963;200:995-6.

18 Matsushima T, Kawabe S, Shibuya M, Sugimura T.Aldolase isozymes in rat tumor cells. Biochem Biophys ResCommun 1968;30:565-70.

19 Tsutsumi K, Tsutsumi R, Ishikawa K. Tissue-specificchanges in chromatin structure of the rat aldolase B locus.JtBiochem 1987;102:1013-21.

20 Zhao Y, Tsutsumi R, Yamaki M, Nagatsuka Y, Ejiri S, Tsut-sumi K. Initiation zone of DNA replication at the aldolaseB locus encompasses transcription promoter region.Nucleic Acids Res 1994;22:5385-90.

21 Daimon M, Tsutsumi K, Ishikawa K. DNA methylation andthe regulation of aldolase B gene expression. J Biochem1986;100: 1279-86.

22 Courtois G, Baumhueter S, Crabtree GR. Purified hepato-cyte nuclear factor 1 interacts with a family of hepatocyte-specific promoters. Proc Natl Acad Sci USA 1988;85:7937-41.

23 Tsutsumi K, Ito K, Ishikawa K. Developmental appearanceof transcription factors that regulate liver-specific expres-sion of the aldolase B gene. Mol Cell Biol 1989;9:4923-3 1.

24 Tsutsumi K, Ito K, Yabuki T, Ishikawa K. AIF-B, a novelCCAAT-binding transcription activator that interacts withthe aldolase B promoter. FEBS Lett 1993;321:51-4.

25 Gregori C, Kahn A, Prichard AL. Activity of the ratliver-specific aldolase B promoter is restrained by HNF3.Nucleic Acids Res 1994;22:1242-6.

26 Ito K, Tsutsumi K, Kuzumaki T, Gomez PF, Otsu K,Ishikawa K. A novel growth inducible gene that encodes aprotein with a conserved cold-shock domain. Nucleic AcidsRes 1994;22:2036-41.

27 Gregori C, Kahn A, Pichard AL. Competition betweentranscription factors HNF1 and HNF3, and alternativecell-specific activation by DBP and C/EBP contribute tothe regulation of the liver-specific aldolase B promoter.Nucleic Acids Res 1993;21:897-903.

28 Sabourin JC, Kern AS, Gregori C, et al. An intronicenhancer essential for tissue-specific expression of thealdolase B transgenes. J Biol Chem 1996;271:3469-73.

29 Peanasky RJ, Lardy HA. Bovine liver aldolase. I. Isolation,crystallisation and some general properties. II. Physical andchemical measurements on the crystalline enzyme. 7 BiolChem 1958;233:365-73.

30 Penhoet E, Kochman M, Valentine R, Rutter WJ. The subu-nit structure of mammalian fructose diphosphate aldolase.Biochemistry 1967;6:2940-9.

31 Horecker BL, Tsolas 0, Lai CY Aldolases. In: Boyer PD,ed. The enzymes. Vol 7. 3rd ed. New York: Academic Press,1972:213-58.

32 Hers HG, Kusaka T. Le metabolisme du fructose-i-phosphate dans le foie. Biochim Biophys Acta 1953;11:427-30.

33 Rutter WJ, Woodfin BM, Blostein RE. Enzymic homology.Structural and catalytic differentiation of fructose diphos-phate aldolase. Acta Chem Scand 1963;17:5226.

34 Rutter WJ. Evolution of aldolase. Fed Proc Fed Am Soc ExpBiol 1964;23: 1248-57.

35 Ingram VM. The hemoglobins in genetics and evolution. NewYork: Columbia University Press, 1963.

36 Tolan DR, Niclas J, Bruce BD, Lebo RV. Evolutionaryimplications of the human aldolase-A, -B, -C, and-pseudogene chromosome locations. Am J Hum Genet1987;41 :907-24.

37 Kawahara K, Tanford C. The number of polypeptide chainsin rabbit muscle aldolase. Biochemistry 1966;5:1578-84.

38 Morse DE, Chan W, Horecker BL. The subunit structureand carboxy-terminal sequence of rabbit muscle aldolase.Proc Natl Acad Sci USA 1967;58:628-34.

39 Sia CL, Horecker BL. The molecular weight of rabbit mus-cle aldolase and the properties of the subunits in acid solu-tion. Arch Biochem Biophys 1968;123:186-94.

40 Sygush J, Beaudry D, Allaire M. Molecular architecture ofrabbit skeletal muscle aldolase at 2.7-A resolution. ProcNatlAcad Sci USA 1987;84:7846-50.

41 Blom N, Sygusch J. Product binding and role of theC-terminal region in class 1 D-fructose 1,6-bisphosphatealdolase. Nature Struct Biol 1997;4:36-9.

42 Banner DW, Bloomer AC, Petsko GA, et al. Structure ofchicken muscle triose phosphate isomerase determinedcrystallographically at 2.5 A resolution. Nature 1975;255:609-14.

43 Alber T, Banner DW, Bloomer AC, et al. On thethree-dimensional structure and catalytic mechanism oftriose phosphate isomerase. Philos Trans R Soc Lond B BiolSci 1981;293:159-71.

44 Stuart DI, Levine M, Muirhead H, Stammers DK. Crystalstructure of cat muscle pyruvate kinase at a resolution of2.6 A. 7Mol Biol 1979;134:109-42.

45 Farber GK, Petsko GA. The evolution of s/p barrelenzymes. Trends Biol Sci1990;15:228-34.

46 Lai CY, Nakai N, Chang D. Amino acid sequence of rabbitmuscle aldolase and the structure of the active center. Sci-ence 1974;183:1204-6.

363



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/

Ali, Rellos, Cox

47 Hartman FC, Brown JB. Affinity labelling on a previouslyundetected essential lysyl residue in class I fructosebisphosphate aldolase. J Biol Chem 1976;251:3057-62.

48 Drechsler ER, Boyer PD, Kowalsky AG. The catalytic activ-ity of carboxy peptidase-degraded aldolase. J Biol Chem1959;234:2627-34.

49 Takahashi I, Takasaki Y, Hori K. Site-directed mutagenesisof human aldolase isozymes: the role of Cys-72 andCys-338 residues of aldolase A and of the carboxy-terminalTyr residues of aldolases A and B.JBiochem 1989;105:281-6.

50 Henry I, Gallano P, Besmond C, et al. The structural genefor aldolase B (ALD B) maps to 9ql3-32. Ann Hum Genet1985;49: 173-80.

51 Lebo RV, Tolan DR, Bruce BD, Cheung MC, Kan YW.Spot-blot analysis of sorted chromosomes assigns afructose intolerance disease locus to chromosome 9.Cytometry 1985;6:478-83.

52 Lench NJ, Telford EA, Andersen SE, Moynihan TP, Robin-son PA, Markham AF. An EST and STS-based YAC con-tig map ofhuman chromosome 9q22.3. Genomics 1996;38:199-205.

53 Tolan DR, Penhoet EE. Characterisation of the humanaldolase B gene. Mol Biol Med 1986;3:245-64.

54 Rottmann WH, Tolan DR, Penhoet EE. Complete aminoacid sequence for human aldolase B derived from cDNAand genomic clones. Proc NatlAcad Sci USA 1984;81:738-42.

55 Santamaria R, Buono P, Paolella G, et al. Aldolase A, aldo-lase B and aldolalse C map to chromosome 16,chromosome 9 and chromosome 17 respectively. CytogenetCell Genet 1987;46:687.

56 Kitajima Y, Takasaki Y, Takahashi I, Hori K. Constructionand properties of active chimeric enzymes between humanaldolases A and B: analysis of molecular regions whichdetermine isozyme-specific functions. J Biol Chem 1990;265:17493-8.

57 Shaw-Lee R, Lissemore JL, Sullivan DT, Tolan DR.Alternative splicing of fructose 1,6-biphosphate aldolasetranscripts in Drosophila melanogaster predicts three iso-zymes. J Biol Chem 1992;267:3959-67.

58 Kai T, Sugimoto Y, Kusakabe T, Zhang R, Koga K, Hori K.Gene structure and multiple mRNA species of Drosophilamelanogaster aldolase generating three isozymes with differ-ent enzymatic properties. J Biochem 1992;1 12:677-88.

59 Zhang R, Kai T, Sugimoto Y, et al. Drosophila melanogasteraldolase: characterisation of the isozymes a, 3 and y gener-ated from a single gene.JBiochem 1995;118:183-8.

60 Hers HG, Joassin G. Anomalie de l'aldolase hepatique dansl'intolerance au fructose. Enzymol Biol Clin 1961;1:4-14.

61 Chambers RA, Pratt RTC. Idiosyncrasy to fructose. Lancet1956;ii:340.

62 Froesch ER, Prader A, Labhart A, Stuber HW, WolfHP. Dieherediatare Fructoseintoleranz, eine bisher nicht bekanntekongentile Stoffwechselstorung. Schweiz Med Wochenschr1957;87:1168-71.

63 Levin B, Snodgrass GJAI, Oberholzer VG, Burgess EA,Dobbs RH. Fructosaemia: observations on seven cases.AmJ Med 1968;45:826-38.

64 Baerlocher K, Gitzelmann R, Steinmann B, Gitzelmann-Cumarasamy N. Hereditary fructose intolerance in earlychildhood: a major diagnostic challenge. Survey of 20symptomatic cases. Helv PaediatrActa 1978;33:465-87.

65 Odievre M, Gentil C, Gautier M, Alagille D. Hereditaryfructose intolerance in childhood: diagnosis, managementand course in 55 patients. Am J Dis Child 1978;132:605-8.

66 Mock DM, Perman JA, Thaler MM, Morris RC. Chronicfructose intoxication after infancy in children with heredi-tary fructose intolerance: a cause of growth retardation. NEnglJ7Med 1983;309:764-70.

67 Ruecker AV, Endres W, Shin YS, Butenandt I, Steinmann B,Gitzelmann R. A case of fatal hereditary fructoseintolerance: misleading information on formula composi-tion. Helv PaediatrActa 1981;36:599-600.

68 Cross NCP, Cox TM. Molecular analysis of aldolase Bgenes in the diagnosis of hereditary fructose intolerance inthe United Kingdom. QJ Med 1989;73:1015-20.

69 Marthaler TM, Froesch ER. Hereditary fructose intoler-ance. Dental status of eight patients. Br Dent J 1967;123:597-9.

70 Newburn E, Hoover C, Mettraux G, GrafH. Comparison ofdietary habits and dental health of subjects with hereditaryfructose intolerance and control subjects. J Am Dent Assoc1980;101:619-26.

71 Saxen L, Jousimies-Somer H, Kaisla A, Kanervo A,Summanen P, Sipila I. Subgingival microflora, dental andperiodontal conditions in patients with hereditary fructoseintolerance. ScandJ3 Dent Res 1989;97: 150-8.

72 Cox TM. Iatrogenic deaths in hereditary fructose intoler-ance. Arch Dis Child 1993;69:413-15.

73 Cox TM. Therapeutic use of fructose: professional freedom,pharmacovigilance and Europe. Q_J Med 1995;88:225-7.

74 Schulte MJ, Lenz W. Fatal sorbitol infusion in patient withfructose-sorbitol intolerance. Lancet 1977;ii:188.

75 Muller-Wiefel DE, Steinmann B, Holm-Hadulla M, WilleL, Scharer K, Gitzelmann R. Infusionsbedingtes Nieren-und Lebervesagen bei undiagnostizierter hereditarerFructose-Intoleranz. Deutsche Med Wochenschr 1983;108:985-9.

76 Steegmanns I, Rittmann M, Bayer JR, Gitzelmann R.Erwachsene mit hereditarer Fructoseintoleranz: Gefahr-dung durch Fructoseinfusion. Deutsche Med Wochenschr1990;115:539-41.

77 Rosien U, Cox TM, Ali M, et al. Akutes hepatorenales Ver-sagen bei hereditarer Fructoseintoleranz. Med Klin 1993;88:553-4.

78 Miller M, Drucker WR, Owens JE, Craig JW, WoodwardHJR. Metabolism of intravenous fructose and glucose innormal and diabetic subjects.JClin Invest 1952;31:115-25.

79 Danks DM, Connellan JM, Solomon JR. Hereditaryfructose intolerance: report of a case and comments on thehazards of fructose infusion. Aust PaediatrJ 1972;8:282-6.

80 Woods HF, Alberti KGMM. Dangers of intravenousfructose. Lancet 1972;ii:1354-7.

81 Perheentupa J, Raivio K. Fructose-induced hyperuricaemia.Lancet 1967;ii:528-31.

82 Bergstrom J, Hultman E, Roch-Norlund AE. Lactic acidaccumulation in connection with fructose infusion. ActaMed Scand 1968;184:359-64.

83 Odievre M, Gautier M, Rieu D. Intolerance hereditaire aufructose du nourisson. Evolution des lesions histologiqueshepatiques sous traitment dietetique prolonge (etude dehuit observations). Arch Fr Pediatr 1969;26:433-43.

84 Cornblath M, Schwartz R. Disorders of carbohydrate metabo-lism in infancy. 3rd ed. Boston: Blackwell ScientificPublications, 1991:325-51.

85 Hers HG. Le metabolisme du fructose. Brussels: Arscia, 1957.86 Cori GT, Ochoa S, Slein MW, Cori CF. The metabolism of

fructose in liver. Isolation of fructose-i-phosphate andinorganic pyrophosphate. Biochim Biophys Acta 1951;7:304-17.

87 Leuthardt F, Testa E. Die Phosphorylierung der Fructose inder Leber. II. Mitteilung. Helv Chim Acta 1951;34:931-8.

88 Hers HG. La fructokinase du foie. Biochim Biophys Acta1952;8:416-23.

89 Froesch ER, Wolf HP, Baitsch H, Prader A, Labhart A.Hereditary fructose intolerance: an inborn defect ofhepaticfructose-i-phosphate splitting aldolase. Am J Med 1963;34:151-67.

90 Pitkanen E, Perheentupa J. Eine biochemische Untersuc-hung uber zwei Falle von Fructoseintoleranz. Ann PaediatrFenn 1962;8:236-44.

91 Cornblath M, Rosenthal IM, Reisner SH, Wybregt SH,Crane RK. Hereditary fructose intolerance. N EnglJ Med1963;269: 1271-8.

92 Landau BR, Marshall JS, Craig JW, Hostetler KY, GenuthSM. Quantitation of the pathways of fructose metabolismin normal and fructose-intolerant subjects. J Lab Clin Med1971;78:608-18.

93 Froesch ER, Ginsberg JL. Fructose metabolism of adiposetissue. I. Comparison of fructose and glucose metabolismin epididymal adipose tissue of normal rats. Y Biol Chem1962;237:3317-24.

94 Woods HF, Eggleston LV, Krebs H. The cause of hepaticaccumulation of fructose-l-phosphate on fructose loading.BiochemJ 1970;119:501-10.

95 Watts RWE, Watts JEM, Seegmiller JE. Xanthine oxidaseactivity in human tissues and its inhibition by allopurinol(4-hydroxypyrazolo (3,4-d) pyrimidine). J7 Lab Clin Med1965;66:688-97.

96 Maenpaa PH, Raivio KO, Kekomaki MP. Liver adeninenucleotides: fructose-induced depletion and its effects onprotein synthesis. Sciencel968;161:1253-4.

97 Bode JC, Zelder 0, Rumpelt HJ, Wittkamp U. Depletion ofliver adenosine phosphates and metabolic effects ofintravenous infusion of fructose or sorbitol in man and inthe rat. EurJ Clin Invest 1973;3:436-41.

98 Van den Berghe G, Bronfman M, Vanneste R, Hers HG.The mechanism of adenosine triphosphate depletion in theliver after a load of fructose: a kinetic study of liveradenylate deaminase. Biochem _J 1977;162:601-9.

99 Oberhaensli RD, Rajagopalan B, Taylor DJ, et al. Study ofhereditary fructose intolerance by use of"P magnetic reso-nance spectroscopy. Lancet 1987;ii:931-4.

100 Maenpaa PH. Fructose-induced alterations in liver poly-some profiles and Mg"' levels. FEBS Lett 1972;24:37-40.

101 Levin B, Oberholzer VG, Snodgrass GJAI, Stimmer L,Wilmers MJ. Fructosaemia: an inborn error of fructosemetabolism. Arch Dis Child 1963;38:220-30.

102 Parks TE, Ben-Gershom E, Lardy HA. Liver fructokinase._JBiol Chem 1957;227:231-42.

103 Smith LH, Ettinger RH, Seligson D. A comparison of themetabolism of fructose and glucose in hepatic disease anddiabetes mellitus. J Clin Invest 1953;32:273-82.

104 Dubois R, Loeb H, Malaisse-Lagae F, Toppet M. Etudeclinique et anatomo-pathologique de deux casd'intolerance congenitale au fructose. Pediatrie 1965;20:5-14.

105 Rossier A, Milhaud G, Colin J, et al. Intolerance congeni-tale au fructose. Deux cas familiaux avec etude bio-chimique in vitro. Arch Fr Pediatr 1966;23:533-52.

106 Kaufmann U, Froesch ER. Inhibition of phosphorylase-aby fructose-I -phosphate, a-glycerophosphate and fructose-1,6-diphosphate: explanation for fructose-induced hy-poglycaemia in hereditary fructose intolerance andfructose-1,6-diphosphatase deficiency. Eur _ Clin Invest1973;3:407-13.

107 Van den Berghe G, Hue L, Hers HG. Effect of the admin-istration of fructose on the glycogenolytic action ofglucagon: an investigation of the pathogeny of hereditaryfructose intolerance. Biochem_J 1973;134:637-45.

108 Thurston JH, Jones EM, Hauhart RE. Decrease and inhi-bition of liver glycogen phosphorylase after fructose: anexperimental model for the study of hereditary fructoseintolerance (HFI). Diabetes 1974;23:597-604.

364



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/


109 Froesch ER, Prader A, Wolf HP, Labhart A. Dieherediatare Fructoseintoleranz. Helv Paediatr Acta 1959;14:99-112.

110 Zalitis J, Oliver IT. Inhibition of glucose phosphate isomer-ase by metabolic intermediates of fructose. Biochem J1967;102:753-9.

111 Eggleston LV, Woods HF. Activation of liver pyruvatekinase by fructose-I-phosphate. FEBS Lett 1970;6:43-5.

112 Andersson G, Brohult J, Sterner G. Increasing metabolicacidosis following fructose infusion in two children. ActaPaediatr Scand 1969;58:301-4.

113 Morris RC Jr. An experimental renal acidification defect inpatients with hereditary fructose intolerance. II. Its distinc-tion from classic renal tubular acidosis; its resemblance tothe renal acidification defect associated with the Fanconisyndrome of children with cystinosis. Jf Clin Invest 1968;47:1648-63.

114 Richardson RMA, Little JA, Patten RL, Goldstein MB,Halperin ML. Pathogenesis of acidosis in hereditaryfructose intolerance. Metabolism: Clin Exp 1979;28: 1133-8.

115 Saxon L, Papper S. Abdominal pain occurring during therapid administration of fructose solutions. N Engl _7 Med1957;256: 132-3.

116 Spaapen LJM. Secondary carbohydrate deficient syn-dromes. Amino Acids 1997;12:390.

117 Cross NCP, Tolan DR, Cox TM. Catalytic deficiency ofhuman aldolase B in hereditary fructose intolerance causedby a common missense mutation. Cell 1988;53:881-5.

118 Cox TM, O'Donnell MW, Camilleri M, Burghes AH. Iso-lation and characterization of a mutant liver aldolase inadult hereditary fructose intolerance. Identification of theenzyme variant by radioassay in tissue biopsy specimens..Clin Invest 1983;72:201-13.

119 James CL, Rellos P, Ali M, Heeley AF, Cox TM. Neonatalscreening for hereditary fructose intolerance: frequency ofthe most common mutant aldolase B allele (A149P) in theBritish population. _7 Med Genet 1996;33:837-4 1.

120 Cross NCP, de Franchis R, Sebastio G, et al. Molecularanalysis of aldolase B genes in hereditary fructoseintolerance. Lancet 1990;335:306-9.

121 Cox TM. Aldolase B and fructose intolerance. FASEB ]1994;8:62-71.

122 Tolan DR, Brooks CC. Molecular analysis of commonaldolase B alleles for hereditary fructose intolerance inNorth Americans. Biochem Med Metab Biol 1992;48: 19-25.

123 Ali M, James CL, Cox TM. A newly identified aldolase Bsplicing mutation (G-*C, 5' intron 5) in hereditary fructoseintolerance from New Zealand. Hum Mutat 1996;7:155-7.

124 Cavalli-Sforza LL, Menozzi P, Piazza A. The history andgeography ofhuman genes. Princeton, New Jersey: PrincetonUniversity Press, 1993.

125 Sebastio G, de Franchis R, Strisciuglio P, et al. Aldolase Bmutations in Italian families affected by hereditary fructoseintolerance. _7 Med Genet 1991;28:241-3.

126 Brooks CC, Tolan DR. Association of the widespreadA149P hereditary fructose intolerance mutation withnewly identified sequence polymorphisms in the aldolase Bgene. Am _7 Hum Genet 1993;52:835-40.

127 Cross NCP, Stojanov LM, Cox TM. A new aldolase Bvariant, N334K, is a common cause of hereditary fructoseintolerance in Yugoslavia. Nucleic Acids Res 1990;18:1925.

128 Dazzo C, Tolan DR. Molecular evidence for compoundheterozygosity in hereditary fructose intolerance. Am _7Hum Genet 1990;46:1194-9.

129 Cox TM, O'Donnell MW, Camilleri M. Allelic heteroge-neity in adult hereditary fructose intolerance: detection ofstructural mutations in the aldolase B molecule. Mol BiolMed 1983;1:393-400.

130 Ali M. The molecular biology of fructose intolerance. PhDThesis, University of Cambridge, UK, 1995.

131 Santamaria R, Scarano MI, Esposito G, Chiandetti L, IzzoP, Salvatore F. The molecular basis of hereditary fructoseintolerance in Italian children. EurJr Clin Biochem 1993;31:675-8.

132 Santamaria R, Tamasi S, Del Piano G, et al. Molecularbasis of hereditary fructose intolerance in Italy: identifica-tion of two novel mutations in the aldolase B gene. _7 MedGenet 1996;33:786-8.

133 Ali M, Cox TM. Diverse mutations in the aldolase B genethat underlie the prevalence of hereditary fructose intoler-ance.Am 7Hum Genet 1995;56:1002-5.

134 Ali M, Tungman G, Cross NCP, et al. Null alleles of thealdolase B gene in patients with hereditary fructose intoler-ance._Med Genet 1994;31:499-503.

135 Brooks CC, Tolan DR. A partially active mutant aldolase Bfrom a patient with hereditary fructose intolerance. FASEB_7 1994;8:107-13.

136 Ali M, Rosien U, Cox TM. DNA diagnosis of fatal fructoseintolerance from archival tissue. Q.7Med 1993;86:25-30.

137 Cross NCP, Cox TM. Partial aldolase B gene deletions inhereditary fructose intolerance. Am .7 Hum Genet 1990,47:101-6.

138 Kajihara S, Mukai T, Arai Y, Owada M, Kitagawa T, HoriK. Hereditary fructose intolerance caused by a nonsensemutation of the aldolase B gene. Am IHum Genet 1990;47:562-7.

139 Nordmann Y, Schapira F, Dreyfus JC. A structurally-modified liver aldolase in fructose intolerance: immuno-logical and kinetic evidence. Biochem Biophys Res Commun1968;31:884-9.

140 Gitzelmann R, Steinmann B, Bally C, Lebherz HG. Anti-body activation of mutant fructose diphosphate aldolase Bin liver extracts of patients with hereditary fructose intoler-ance. Biochem Biophys Res Commun 1974;59:1270-7.

141 Lamiere N., Mussche M, Baele G, Kint JA, Ringoir S.Hereditary fructose intolerance: a difficult diagnosis in theadult. AmJMed 1978;65:416-23.

142 Schapira F, Hatzfeld A, Gregori C. Studies on liveraldolases in hereditary fructose intolerance. Enzyme 1974;18:73-83.

143 Gregori C, Shapira F, Kahn A, Delpech, Dreyfus JC.Molecular studies of liver aldolase B in hereditary fructoseintolerance using blotting and immunochemical tech-niques. Ann Hum Genet 1982;46:281-92.

144 Sakakibara M, Takahashi T, Takahashi Y, Mukai T, HoriK. Construction and expression ofhuman aldolase A and Bexpression plasmids in Escherichia coli host. BiochemBiophys Acta 1989;1007:334-42.

145 Gamblin SJ, Davies GJ, Grimes JM, Jackson RM,Littlechild JA, Watson HC. Activity and specificity ofhuman aldolases. .7 Mol Biol 1991 ;219:573-6.

146 Morris AJ, Tolan DR. Lysine- 146 of rabbit muscle aldolaseis essential for cleavage and condensation of the C3-C4bond of fructose 1,6-bis(phosphate). Biochemistry 1994;33:12291-7.

147 Berthiaume L, Tolan DR, Sygusch J. Differential usage ofthe carboxyl-terminal region among aldolase isozymes. .Biol Chem 1993;268:10826-35.

148 Motoki K, Kitajima Y, Katsuji H. Isozyme-specificmodules on human aldolase A molecule. .7 Biol Chem 1993;268:1677-83.

149 Ali M, Sebastio G, Cox TM. Identification of a novelmutation (Leu256"-Pro) in the human aldolase B gene asso-ciated with hereditary fructose intolerance. Hum Mol Genet1994;3:203-4.

150 Beernink PT, Tolan DR. Subunit interface mutants of rab-bit muscle aldolase form active dimers. Protein Sci 1994;3:1383-91.

151 Beernink PT, Tolan DR. Disruption of the aldolase Atetramer into catalytically active monomers. Proc NadAcadSci USA 1996;93:5374-9.

152 Cox TM, Camilleri M, O'Donnell MW, Chadwick VS.Pseudodominant transmission of fructose intolerance in anadult and three offspring. N EnglJ Med 1982;307:537-40.

153 Steinmann B, Gitzelmann R. The diagnosis of hereditaryfructose intolerance. Helv Paediatr Acta 1981;36:297-316.

154 Saiki RK, Scharf S, Faloona F, et al. Enzymatic amplifica-tion of ,-globin genomic sequences and restriction siteanalysis for diagnosis of sickle cell anemia. Science1985;230:1350-4.

155 Gitzelmann R, Baerlocher K. Vorteile und Nachteile derFructose in der Nahrung. PadiatrFortbild Prax 1973;37:40-55.

156 Garrod AE. The incidence of alkaptonuria: a study inchemical individuality. Lancet 1902;ii: 1616-20.

157 Wolf H, Zschocke D, Wedemeyer FW, Hubner W.Angeborene hereditare fructose-intoleranz. Klin Wochenschr1959;37:693-6.

158 Linden L, Nisell J. Hereditar fructosintolerans. SvenskaLakartidn 1964;61:3185-95.

159 Barry RGG, St Colum SR, Magner JW. Hereditaryfructose intolerance in parent and child. _7 Irish Med Assoc1968;61:308-10.

160 Rampa M, Froesch ER. Eleven cases of hereditary fructoseintolerance in one Swiss family with a pair of monozygoticand of dizygotic twins. Helv Paediatr Acta 1981;36:317-24.

161 Heine W, Schill H, Tessmann D, Kupatz H. Letale Leber-dystrophie bei drei Geschwistern mit hereditarer Frukto-seintoleranz nach Dauertropfinfusionen mit sorbitolhalti-gen Infusionslosungen. Deutsche Gesundheitswesen 1969;24:2325-9.

162 Hackl JM, Balogh D, Kunz F, et al. Postoperative Frukto-seinfusion bei wahrscheinlich hereditarer Fruktoseintoler-anz. Wien Klin Wochenschr 1978;90:237-40.

163 Rey M, Behrens R, Zeilinger G. Fatale Folgen einerFructose-Infusion bei undiagnostizierter Fructose-Intoleranz. Deutsche Med Wochenschr 1988;1 13:945-7.

164 Brooks CC, Buist N, Tuerck J, Tolan DR. Identification ofa splice-site mutation in the aldolase B gene from an indi-vidual with hereditary fructose intolerance. Am .7 HumGenet 1991;49:1075-81.

365



g.bmj.com

/J M


g.35.5.353 on 1 May 1998. D

ownloaded from

http://jmg.bmj.com/

review article hereditary fructose intolerancehereditary fructose intolerance (hfi, omim 22960) is a...

Documents