Med Microbiol Immunol (1995) 184:139-146 �9 Springer-Verlag 1995

C a r s t e n F r a n k �9 K e r s t i n S t e i n e r �9 H o r s t M a l k e

Conservation of the organization of the streptokinase gene region among pathogenic streptococci

Received: 20 July 1995

A b s t r a c t Using ten gene-specific probes from the cloned and sequenced streptokinase gene (skc) region (8 931 bp) of Streptococcus equisimiIis H46A, a human serogroup C strain, the conservation of these genes and their linkage re- lationships were studied by Southern hybridization in path- ogenic streptococci differing taxonomically, serologically, in regard to their host range, and in the class of plasmino- gen activator produced. The results indicate that in S. pyo- genes (strains A374, NZ131 and SF130/13) and a human group G strain (G19 908) both gene content and gene or- der as determined for H46A (dexB-abc-lrp-skc-orfl-rel) are preserved. The same is true of an equine S. equisimilis isolate (87-542-W), the streptokinase gene of which has been shown to hybridize detectably with skc, a result at variance with that obtained previously by others. In con- trast, the chromosomal DNA of three S. uberis strains (0140J, C198, C216) of bovine origin, two of which pro- duced a plasminogen activator different from streptoki- nase, hybridized only with dexB-, abc- and rel-specific probes, and the homologues of these genes appeared to lie close to each other. The maintenance of the organization of the streptokinase gene region in strains differing in over- all chromosomal character suggests that this gene arrange- ment is of selective advantage.

Key words Streptococcus equisimilis �9 S. pyogenes �9 S. uberis �9 Streptokinase gene region - Southern hybridization �9 Genome organization

Introduction

The streptokinase gene is a virulence determinant present in virtually all human isolates of serogroup A, C and G streptococci [7]. Recent sequencing of the chromosomal

C. Frank �9 K. Steiner. H. Malke (t~) Institute for Molecular Biology, Jena University, Winzerlaer Strasse 10, D-07745 Jena, Germany Tel.: 49-3641-657530; Fax: 49-3641-657520; e-mail: hmalke @imb-jena.de

streptokinase gene (skc) region of Streptococcus equisim- ilis H46A, a human group C strain used for commercial streptokinase production, has identified five genes or open reading frames (ORFs) adjacent to skc [14]. Contained in a DNA segment of 8931 base pairs (bp), these genes are encoded on the opposite strand to skc and are arranged as follows: dexB-abc-lrp-skc-orfl-rel. The dexB gene codes for an intracellular a-glucosidase, and abc encodes an ATP-binding cassette (ABC) transporter protein homolo- gous to sugar or sugar-related transporters of gram-posi- tive and gram-negative bacteria. The lrp gene adjacent to, and transcribed divergently from, skc codes for a leucine- rich protein of unknown function that has a leucine zipper motif at its C-terminal end. The rel gene is a structural ho- mologue of the Escherichia coli homologous relA and spoT genes that function in the synthesis and degradation of gua- nosine 5',3' polyphosphate [(p)ppGpp] during the stress of nutrient limitation and during entry into the stationary phase [2, 5]. Finally, orfl, of unknown function and co- transcribed with rel, is homologous to the E. coli orf145 gene (EMBL accession number L 19 201), the function of which is also not known.

To date, the primary structures of six streptokinases from different group A, C and G strains have been com- pletely determined, five being deduced from the nucleo- tide sequence of their genes [3, 8, 12, 24, 25] and one de- termined by chemical sequencing [ 18]. Sequence compar- ison revealed that the proteins are heterogeneous, with the overall sequence identity between any pair of molecules ranging from 80% to 98%. An unrooted phylogenetic tree for the six proteins identified two primary branches [13]. One branch contains the closely related streptokinases from some serogroup A, C and G strains, including H46A. The other major lineage comprises, of the completely se- quenced proteins, only two group A streptokinase asso- ciated with nephritogenic strains. The allocation of the whole molecules to one or the other main cluster coincides with the primary branching of their internal 72-amino acid variable region, V1, showing that allelic variation mani- fests itself predominantly in the central region of the genes. A subsequent molecular population genetic analysis of the

s t rep tokinase gene f rom group A s t reptococci inc luded the var iab le regions f rom 47 isola tes and conf i rmed the exis- tence o f two major l ineages , which, however , showed no s imple re la t ionship be tween s t reptokinase al lele and spe- cif ic d isease [9].

Given the ex is tence of at least two major classes of s t rep tokinase and the occur rence o f c lose ly re la ted strep- tok inase genes in s t rep tococca l species that are wel l dif- ferent ia ted in overa l l c h r o m o s o m a l character , the quest ion arises as to whether or not the genomic envi ronment o f the s t rep tokinase gene locus has been conserved in different pa thogenic s t rep tococca l strains. This quest ion is of con- s iderable in teres t because the order of genes inf luences not only l inkage but also gene regulat ion. We approached this p r ob l em using gene-spec i f i c probes f rom the s t reptokinase gene region o f H 4 6 A in Southern hybr id iza t ion exper i - ments , inc luding the ch romosoma l D N A from four human s t rep tococca l i sola tes which produce s t reptokinases that are represen ta t ive o f the two majo r classes. In addit ion, three ve te r inary s t rep tococca l strains p roduc ing p la smino- gen ac t iva tors that are more or less specif ic for the plas- minogens o f their homologous hosts were inc luded in this study.

Materials and methods

Streptococcal strains

The properties of the streptococcal strains investigated are listed in Table 1.

Gene-specific probes

DNA probes were isolated from an Escherichia coli plasmid library covering the 8931-bp segment of the S. equisimilis H46A skc gene region (Table 2). Recombinant plasmid DNA was prepared by sol- id-phase anion-exchange separation as recommended by the suppli- er of the QIAGEN Plasmid Kit [19]. Restricted plasmid DNA was subjected to agarose gel electrophoresis, and appropriate restriction fragments to be used as probes were purified with the QIAEX Gel Extraction Kit [20]. The probes (0.5 gg of DNA) were labeled with [a-35S]dATP (37 TBq/mmol) using a random primed DNA labeling kit (Boehringer), following the manufacturer's protocoll. Unincor-

porated nucleotides were removed from the reaction mixtures by Se- phadex G50 chromatography. The properties of the probes used are specified in Table 2 (see also Fig. 2).

Southern hybridization

Genomic streptococcal DNA was isolated [1] from 50-ml cultures grown overnight at 37 ~ without agitation, in brain-heart infusion medium (Difco) supplemented with 20 mM glycine, 30 mM glucose and 2% horse serum. An estimated 5 ~tg of appropriately restrict- ed DNA [22] was loaded per lane on agarose gels (0.8-1%). Follow- ing electrophoresis, gels were soaked for 8 min in each of 0.25 N HC1, 0.5 M NaOH-1.5 M NaCI, and 0.5 M TRIS-1.5 M NaC1 (pH 7.4). The DNA was then transferred by vacuum blotting to Hy- bond-N nylon membranes (Amersham). After fixation of the DNA by UV cross-linking and drying at 80 ~ the blots were prehybri- dized in Hepes buffer-10x Denhardt's reagent-herring ~perm DNA [22]. Hybridizations with ca. 100 ng of probe per 100 cm membrane area were performed overnight at 65 ~ Blots were washed twice for 20 min in 2x SSC-0.1% SDS at 60 ~ before being autoradiographed. Membranes to be used for rehybridization with a different probe were treated as detailed by Sambrook et al. [22] to remove old probe DNA.

Data recording

In each gel, phage lambda DNA digested with HindIIl, EcoRI, or a combination of these enzymes, served as the migration standard. Au- toradiography was performed with the FUJIX BAS 1000 Bio-Imag- ing Analyzer System (Fuji Photo Film, Japan), using the TINA 2.07c software from Raytest (Isotopenmegger~ite GmbH, Straubenhardt, Germany).

Results

To de termine the conservat ion and the ar rangement of the six genes ident i f ied in the skc region o f H 4 6 A among the test strains, D N A probes were chosen such as to consis t of an internal coding region of each of the six genes or to span two or more adjacent genes o f the region (Table 2 and see Fig. 2). In some cases, double hybr id iza t ions with probes internal to two adjacent H46A genes were pe r fo rmed to c lar i fy l inkage relat ionships. Of the enzymes used to c leave the target D N A s including H46A D N A as reference stan- dard, NdeI and Bcl I produc ing large fragments , and

Table 1 Properties of the streptococcal strains used a

Strain Serogroup Origin Disease Plasminogen activator Ref. (M type)

S. pyogenes A374 A (12) Human APSGN b S. pyogenes NZ131 A (49) Human APSGN S. pyogenes SF130/13 A (1) Human Scarlet fever c S. equisimilis H46A C Human General c S. equisimilis 87-542-W C Equine G19908 G Human Urethritis c S. uberis 0140J Bovine Mastiffs S. uberis C216 Bovine S. uberis C198 Bovine

Streptokinase; protein sequence known [18] Streptokinase; gene/protein sequence known [8] Streptokinase; gene/protein sequence known [24] Streptokinase; gene/protein sequence known [12] Streptokinase; sequence unknown [17] Streptokinase; gene/protein sequence known [25] Plasminogen activator different from streptokinase; sequence unknown [10] Plasminogen activator different from streptokinase; sequence unknown [10] No plasminogen activator detected (control strain) [10]

a Blank, no information available b Acute poststreptococcal glomerulonephritis c Not associated with APSGN

Table 2 Characteristics of the DNA probes used

141

Probe a Coordinates b Size (bp) %(G+C) Tm(~ Plasmid Ref. (restriction source fragment)

dexB* 833-1428 595 45.9 95.6 pSHDI4 [14] ( EcoRI/SphI)

abc* 2017-2581 564 44.1 94.8 pUS2 c (PstI/EcoRI)

lrp* 3346-3738 392 45.7 95.0 pMF1 [11] (HindlII/SalI)

skc* 4635 -5265 630 41.1 93.7 pMF 1 [ 11 ] ( ClaI/HindlII)

orfl* 5858-6189 331 42.6 93.5 pSU31 [23] (SpeI/PstI)

rel* 6690-7813 1123 45.0 95.7 pMF 1 [ 11 ] (BamHI/SalI)

dexB-abc* 1-2455 2454 43.6 9 5 . 5 pSHD14 [t4] (HindlII)

dexB-abc-lrp* 1428-3738 2310 42.3 94.9 pUS2 c (SphI/SalI)

lrp-skc-orfl* 3625-6189 2564 39.5 93.8 pMM191 [16] (PstI)

lrp-skc-orfl-rel* 3738-6690 2952 40.1 94.1 pSU31 [23] ( SalI/BamHI)

a Probes are marked with asteriks to distinguish them from corresponding gene symbols b Probe coordinates relate to the numbering of the 8931-bp segment of the skc region from H46A ([ 14]; EMBL accession number, X72 832) r pUC19 derivative carrying an SphI/SalI fragment containing the complete abc gene and p~tial se- quences from dexB and lrp (our unpublished result)

HindIII alone or a combination of HindlII plus BclI yield- ing relatively small fragments proved to be most useful to obtain conclusive and internally consistent results.

Characterization of the streptokinase gene region of human group A and group G strains

Of the human isolates investigated, the nephritogenic strains A374 and NZ131 produce streptokinases which be- long to one of the two major branches of the phylogenetic tree, and the streptokinases of SF130/13, G19908 and H46A belong to the other main lineage [ 13]. The restricted DNAs of all group A and G strains gave intense hybridiza- tion signals with each of the probes used, indicating the presence of structural homologues to all genes of the H46A skc region. The number and size of the various restriction fragments detected with the different probes are summar- ized in Table 3. Evaluation of the various hybridization pat- terns found allows the conclusion that not only the indi- vidual genes but also their linkage relationships and orien- tations relative to the streptokinase gene are completely conserved. Thus, for example, all six gene-internal probes detected a single NdeI restriction fragment of strain-spe- cific size in four of the strains. The DNA of the remaining strain, NZ 131, showed two hybridizing NdeI fragments, an 11.0-kb fragment containing dexB and abc, and a 5.6-kb

fragment containing the lrp-rel segment. The contiguous arrangement of these two fragments was then deduced from the detection of one specific 2.9-kb HindlII fragment that hybridized with abc*, lrp*, as well as ske* (Table 3). By the same token, all six genes were deduced to be contained in two linked Bcll fragments of strain-specific size (Fig. 1). In other cases, adjacency of two test strain genes was ap- parent when a mixture of two gene-internal probes repre- sentative of adjacent genes in H46A detected no fragments other than those also found in hybridization experiments with the separated probes. For this purpose, mixtures of abc*+Irp* and lrp*+skc* were used to establish the con- servation of the abc-lrp-skc segment in all strains. Further- more, the hybridization patterns of the HindlII-digests probed with gene-internal DNAs were consistent with those obtained with probes spanning several H46A genes (Table 3). The same was true of hybridization patterns ob- tained for target DNAs double digested with HindlII and BclI when compared to the patterns obtained after single digestion with these two enzymes. As result of such a bot- tom-up approach, the physical-genetic maps of the strep- tokinase gene region of the test strains are presented in Fig. 2. These maps also show that there exists extensive BclI and HindlII restriction fragment length polymor- phism, with only one specific HindlII site in each the strep- tokinase gene and probably the abc gene, and one BclI site in the lrp gene being conserved throughout.

142

Table 3 Number and size (kb) of DNA restriction fragments from five human streptococcal isolates that hybridize with gene probes iso- lated from the streptokinase gene (skc) region of H46A a

Strain Probe

dexB* abc* lrp* skc* orfl* rel* dexB-abc* dexB-abc- lrp-skc- lrp-skc- lrp* orfl * orfl -rel*

NdeI cleavage A374 7.8 7.8 7.8 7.8 7.8 7.8 NZ131 11.0 11.0 5.6 5.6 5.6 5.6 H46A 7.6 7.6 7.6 7.6 7.6 7.6 SF130/13 11.5 11.5 11.5 11.5 11.5 11.5 G19908 7.9 7.9 7.9 7.9 7.9 7.9

BclI cleavage A374 3.8 3.8 3.8 6.5 6.5 6.5 NZ131 3.9 3.9 3.9 6.5 6.5 6.5 H46A 3.6 3.6 3.6 3.9 3.9 3.9 SF130/13 3.9 3.9 3.9 6.0 6.0 6.0 G19908 5.0 5.0 5.0 4.0 4.0 4.0

HindlII cleavage A374 1.7/0.8 2.9/0.5 2.9 2.9 5.2 5.2 NZ131 1.7/0.8 2.9/0.5 2.9 2.9 2.7 2.7 H46A 2.5 2.5/0.5 1.9 1.9 2.6 2.6 SF130/13 2.4 2.9/0.5 2.9 2.9 2.2 2.2 G19908 4.9 4 .9/0 .5 2.0 2.0 2.7 2.7

HindlII/BclI cleavage A374 1.0/0.8 1.4/0.5 1.4 5.2 NZ131 1.0/0.8 1.4/0.5 1.4 2.7 H46A 2.2 2.2 1.4 2.5 SF130/13 1.7 1.4/0.5 1.4 2.2 G19908 3.6 3.6 1.4 2.5

1.7/0.8/0.5 2.9/0.8/0.5 5.2/2.9 5.2/2.9 1.7/0.8/0.5 2.9/0.8/0.5 2.9/2.7 2.9/2.7 2.5 2.5/1.9/0.5 2.6/1.9 2.6/1.9 2 .4 /0 .5 2 .9 /2 .4 /0 .5 2.9/2.2/0.5 2.9/2.2/0.5 4.9 4.9/2.0/0.5

a Blank, not done

Southern hybridization of DNA from the veterinary isolates

Of the plasminogen activators produced by the streptococ- cal strains of equine or bovine origin included in this study, only the protein elaborated by strain 87-542-W can be con- sidered a streptokinase as judged by its molecular weight (M r -49 000) and the limited sequence identity detected by comparison of the sequenced N terminus of the 87-542-W protein with the N termini of the present group A, C and G streptokinase [ 17]. The plasminogen activator produced by strain 0140J has a molecular weight (M r -29 000) dif- ferent from that of streptokinase (M r -47 000) and appears to exist as a dimeric molecule [10]. Of the remaining S. uberis strains, C216 is known to produce a plasminogen activator which, however, has not yet been characterized, and C198 was found to elaborate no plasminogen activa- tor (W. Ditcham, personal communication, 1994).

The genomic DNA of the four veterinary isolates was digested with HindlII or BclI, and, in the case of 87-542-W, double digests were also investigated. The restricted DNAs were probed with the H46A driver DNAs characterized above (Table 2). Strain 87-542-W DNA gave distinct hy- bridization signals with all probes used (Table 4). Of par-

ticular interest is the detection of both HindlII fragments (1.7 and 1.6 kb) and BclI fragments (6.5 and 2.5 kb) that hybridized with skc*, albeit less intensely than the frag- ments identified with dexB*, abc*, orfl* and rel* (Fig. 3). This result contrasts with the finding of Nowicki et al. [17] who, using an even larger skc probe under formamide hy- bridization conditions, failed to detect homology of skc with any of the equine isolates, including 87-542-W. The discrepancy may be explained by radio-imaging being more sensitive to detect weak hybridizations than other techniques. (As a matter of fact, when testing blots to be rehybridized for the removal of old probe DNA by radio- imaging, we usually detected traces of remaining radioac- tivity that escaped detection even after prolonged exposure to X-ray films; see, for instance, Fig. 1.) Concerning the linkage relationships of the genes in the streptokinase gene region of 87-542-W, the HindlII-digested target DNA al- lowed no firm conclusion because the largest fragments detected by most of the probes (except orfl*) had sizes of only 1.7-1.6 kb. However, dexB*, abc*, lrp* and skc* de- tected a specific BclI fragment of 6.5 kb, and orfl* and rel* hybridized with a 2.5-kb BclI fragment which was also detected with skc* (Fig. 3). In addition, rel* also hybri- dized with a 9.5-kb Bcl fragment. Thus, skc* links the

Fig. 1 Southern hybridization of BclI-digested chromosomal DNA from human group A, C and G isolates with the indicat- ed probes. For each strain, dexB*, abc* and lrp* detect one and the same strain-specif- ic BclI fragment, and, similarly, skc*, orfl* and rel* detect an- other strain-specific fragment. Further, since in all strains one and the same strain-specific HindlII fragment hybridizes with both lrp* and skc* (see Table 3), the above BclI frag- ments containing either dexB, abc and b;o, or skc, orfl and rel must be contiguous. Note that the blots on the bottom were re- hybridized and in some cases show traces of the probe used at first (top row)

143

6.5-kb and 2.5-kb BclI fragments, and rel* links the 2.5-kb and 9.5-kb BcII fragments, indicating that the gene order in strain 87-542-W is similar, if not identical, to that in H46A.

The DNAs from the three S. uberis strains gave hybrid- ization signals with abc* and tel* after cleavage with ei- ther HindlII or BclI, whereas dexB* hybridized only with DNAs digested with HindlII (Table 4). Consistent with this observation, the HindIII-digested DNAs also hybridized with the composite probes provided they carried dexB-, abc- or tel-specific sequences. None of the S. uberis DNAs hybridized with lrp*, skc*, orfl *, or lrp-skc-orfl *. Regard- ing skc and strains 0140J and C216, the same observation was made elsewhere with a digoxigenin-labelled skc se- quence amplified by PCR (W. Ditcham, personal commu- nication, 1994). Taken together, the hybridization patterns (Table 4) suggest that the homologues of dexB and abc are contained in one strain-specific HindlII fragment and do not rule out the possibility that the abc and rel homologues of the S. uberis strains are contained in one strain-specific

BclI fragment. However, further work is needed to estab- lish the linkage relationships of these genes.

Discussion

Using comparative Southern hybridization analysis, we ap- proached questions about the structural maintenance of the streptokinase gene region in pathogenic streptococci dif- fering taxonomically, serologically, in regard to their host range, and in the class of plasminogen activator produced. The results show that in group A streptococci and human isolates of group C and G streptococci the genes of the streptokinase region and their linkage relationships as es- tablished for S. equisimilis H46A [14] are highly con- served. The same appears to be true of an S. equisimilis strain (87-542-W) of equine origin which produces a strep- tokinase that preferentially activates the plasminogen of its homologous host (ESK) and differs structurally and

144

A374

I I I I I I I H B H HH B H

II BH

NZ131

SF130/13

G19908 I I H B

H46A

I I H B

I H

I I HH

I I I B H H

| /rp' ska 'off1

I I I I I H B HH B

I I I I HHH B

II I1: : : HB H HHH B

! ! ! dexB abc /rp

I 81 I 71 t

I 1 ' I I HH H B

ska

I II H BH

skg

' II I H BH

skc off/ re/

10 I

91

I B

I I1 kb

Fig. 2 Physical-genetic maps of the streptokinase gene regions of group A (A374, NZ131, SF130/13), group C (H46A) and group G (G 19 908) streptococci of human origin as deduced from the results of comparative Southern hybridization analyses (see Table 3). The coding regions and orientations of genes with known sequences are indicated by bold arrows below the restriction maps. Thin bars orig- inating from the genes represent sequenced intergenic regions or the truncated coding sequences of the genes flanking the NZ131 ska gene. The position of the gene probes (1, dexB*; 2, abc*; 3, lrp*, 4, skc*; 5, orfl*; 6, rel*; 7, dexB-abc*; 8, dexB-abc-lrp*; 9, lrp-skc- orfl*; 10, lrp-skc-orfl-rel*) specified in Table 2 are also indicated. B and H indicate sites for BclI and HindlII, respectively, with bold capitals representing sites conserved throughout

antigenically f rom the streptokinases elaborated by human strains (HSK) [17]. Given the earlier failure to detect D N A hybridization between the streptokinase gene of 87-542-W and skc [17], the question arose o f whether or not ESKs and HSKs are homologous proteins. Our results show clearly that, provided hybridizat ion signals are visualized with sensitive techniques, these two genes can be demon- strated to hybridize, at least to some extent. This finding, taken together with the observation that the streptokinase gene o f 87-542-W and skc lie in similar genomic environ- ments, leaves no doubt that ESKs and HSKs are geneti- cally homologous . Knowledge of the restriction fragments o f 87-542-W that hybridize with probes f rom the strepto- kinase gene region of H46A (Table 4) should facilitate fu- ture efforts to clone an ESK gene and, ultimately, deter- mine its pr imary structure.

In contrast to the ESK gene, the plasminogen activator genes o f the S. uberis strains o f bovine origin are not ho- mologous to skc, and this is supported by some properties

Fig. 3 Southern hybridization of HindlII- and BclI-digested chro- mosomal DNA from strain 87-542-W with the indicated probes. Note the occurrence of a number of faint signals to which we attach no significance (compare with Table 4). Thus, for instance, in BclI-di- gested DNA orfl* and rel* produce weak bands at -6.5 kb that do not identify genuine fragments. Similarly, the faint signal at -0.8 kb produced in HindlII-digested DNA by skc* is not a significant hybrid- ization signal because a probe of 630 bp (see Table 2) cannot detect three different linked fragments, each of which is larger than the probe

of the protein isolated from strain 0140J [10]. Apparently, this plasminogen activator has no relationship to strepto- kinase, a situation which is reminiscent of streptokinase and staphylokinase which show no homology either [12]. Nevertheless, o f the genes of the streptokinase region, abc, rel and probably dexB homologues can be found in S. ube- ris, and there are some indications that they lie close to each other (Table 4). From an evolutionary standpoint, it will be interesting to learn whether or not the plasminogen activator gene of S. uberis maps in this region.

145

Table 4 Number and size (kb) of DNA restriction fragments from four veterinary streptococcal isolates that hybridize with gene probes from the streptokinase gene (skc) region of H46A a

Strain Probe

dexB* abc* lrp* skc* off1* rel* dexB- dexB- lrp-skc- lrp-skc- abc* abc-lrp* orfl* orfl-rel*

HindIII cleavage 87-542-W 1.7 1.7/0.5 1.6 1.7/1.6 0.7 1.6 1.7/0.8 1 .7 /0 .5 1 .6 /0 .7 1.6/0.7 C198 3.6 3.6 b _ _ 5.4 3.6/0.9 3.6 - 5.4 C216 3.7 3.7 - - - 3.9 3.7 3.7 - 3.9 0140J 3.5 3.5 - - - 3.7 3.5 3.5 - 3.7

BclI cleavage 87-542-W 6.5 6.5 6.5 6.5/2.5 2.5 9.5/2.5 C198 - 5.9 - - 5.9 C216 - 7.4 - - 7.4 0140J - 7.3 - - 7.3

HindlII/BclI cleavage 87-542-W 1.7 1.7/0.5 1.6 1.7/1.4 0.7 1.4 1.7/0.8 1.7/0.5 0.7 1.4/0.7

a Blank, not done b --, no hybridization

With regard to the nature of the genes, the skc region shows some important peculiarities. Apparently, the strep- tokinase gene is the only virulence gene in the sequenced region. It is abundantly and independently transcribed as monocistronic mRNA [14] from a complex promoter which, for full activity, depends on an upstream region rich in AT tracts [4]. The available sequence information from NZ131 [8] and H46A [12] shows that sequence elements involved in the expression of the streptokinase gene [4] are highly conserved in these strains. Moreover, the general or- ganization of the genomic environment of the streptokinase locus suggests that this sequence arrangement is of selec- tive advantage. Since there exists molecular epidemiolog- ical evidence that the streptokinase alleles have a mosaic structure, and horizontal transfer and recombination may be involved in their evolution [9], it seems important to note that each allele occupies a single position relative to other genes in the region. Thus, allele exchanges from one site or one orientation to another by recombinational events may be at a selective disadvantage, if they do occur at all.

From a regulatory point of view, the presence of the abc locus in the streptokinase gene region is hard to under- stand. In S. mutans, which does not carry a streptokinase gene, the abc homologue, msmK, is part of the msm operon involved in the uptake and metabolism of multiple sugars [21]. Furthermore, the known organization of many other prokaryotic transport operons shows the ABC transporter locus tO be generally associated with functionally related genes, in particular those encoding solute binding proteins [6]. Thus, the original finding [14] that, in H46A, the abc gene is detached from functionally related genes and, in addition, shows transcriptional independence to an appre- ciable extent, might have been regarded merely as a freak of nature without widespread occurrence. Although the present experiments establish that, in streptococci carry- ing a streptokinase locus, abc is generally separated from

genes of related function, the functional significance of the dispersal as well as the genetic mechanism that caused it remain elusive.

Previous colony hybridization experiments have shown that the lrp gene is present only in strains that also carry a streptokinase gene [ 14]. The present experiments confirm and expand this observation in establishing the fixed posi- tion of the Irp locus relative to the other loci of the region. Since mutants in which Irp is insertionally inactivated have no readily recognizable phenotype [ 14], and the data banks contain no entries with primary structure similarities to lrp or its product, the function of this gene remains to be elu- cidated.

Of general importance is the identification of rel gene homologues in all strains studied. At present, sequence in- formation and functional analysis of genes homologous to relA and spoT from E. coli are scarce and, particularly for gram-positive pathogenic bacteria, highly desirable. In H46A, the rel gene product mediates the degradation and synthesis of (p)ppGpp [15], a regulatory signal molecule involved in the global control of gene expression [2]. Given our ignorance of the possible role of ppGpp in controlling virulence traits, it will be interesting to continue exploring the streptococcal rel locus and its involvement in adaptive responses to environmental changes.

Acknowledgements We thank W. Ditcham, Institute for Animal Health, Compton, UK, and R. Lottenberg, University of Florida, Gainesville, Fla., for providing bacterial strains. This work was sup- ported by grants from Deutsche Forschungsgemeinschaft (Ma 1330- 1/2) and Fonds der Chemischen Industrie (400 126) to H.M.

References

1. Caparon MG, Scott JR (1991) Genetic manipulation of patho- genic streptococci. Methods Enzymol 204:556-586

146

2. Cashel M, Rudd KE (1987) The stringent response. In: Neid- hardt FC (ed) Escherichia coli and Salmonella typhimurium. American Society for Microbiology, Washington, DC, pp 1410-1438

3. Estrada MR Hernandez L, Perez A, Rodriguez P, Serrano R, Ru- biera R, Pedraza A, Padron G, Antuch W, de la Fuente J, Her- rera L (1992) High level expression of streptokinase in Esche- richia coli. Biotechnology 10:1138-1142

4. Gase K, Ellinger T, Malke H (1995) Complex transcriptional control of the streptokinase gene of Streptococcus equisimilis H46A. Mol Gen Genet 247:749-758

5. Gentry DR, Hernandez V J, Nguyen LH, Jensen DB, Cashel M (1993) Synthesis of the stationary-phase sigma factor o ~ is pos- itively regulated by ppGpp. J Bacteriol 175:7982-7989

6. Higgins CF (1992) ABC transporters: from microorganisms to man. Annu Rev Cell Biol 8:67-113

7. Huang TT, Malke H, Ferretti JJ (1989) Heterogeneity of the streptokinase gene in group A streptococci. Infect lmmun 57:502-506

8. Huang TT, Malke H, Ferretti JJ (1989) The streptokinase gene of group A streptococci: cloning, expression in Escherichia co- li, and sequence analysis. Mol Microbiol 3:197-205

9. Kapur V, Kanjilal S, Hamrick MR, Li L-L, Whittam TS, Saw- yer SA, Musser JM (1995) Molecular population genetic anal- ysis of the streptokinase gene of Streptococcus pyogenes: mo- saic alleles generated by recombination. Mol Microbiol 16:509-519

10. Leigh JA (1994) Purification of a plasminogen activator from Streptococcus uberis. FEMS Microbiol Lett 118:153-158

11. Malke H, Ferretti JJ (1984) Streptokinase: cloning, expression, and excretion by Escherichia coli. Proc Natl Acad Sci USA 81:3557-3561

12. Malke H, Roe B, Ferretti JJ (1985) Nucleotide sequence of the streptokinase gene from Streptococcus equisimilis H46A. Gene 34:357-362

13. Malke H, Steiner K, Gase K, Mechold U, Ellinger T (1995) The streptokinase gene: allelic variation, genomic environment, and expression control. In: Ferretti JJ, Gilmore MS, Klaenhammer TR, Brown F (eds) Genetics of streptococci, enterococci and lac- tococci. Karger, Basel, pp 183-193

14. Mechold U, Steiner K, Vettermann S, Malke H (1993) Genetic organization of the streptokinase region of the Streptococcus equisimilis H46A chromosome. Mol Gen Genet 241:129-140

15. Mechold U, Cashel M, Steiner K, Gentry D, Malke H (1994) Functional analysis of the rel gene of Streptococcus equisimilis. 4th International ASM Conference on Streptococcal Genetics, Santa Fe, N.M. American Society for Microbiology, Washing- ton, DC, abstract T8

16. Mtiller J, Reinert H, Malke H (1989) Streptokinase mutations relieving Escherichia coli K- 12 (prlA4) of detriments caused by the wild-type skc gene. J Bacteriol 171:2202-2208

17. Nowicki ST, Minning-Wenz D, Johnston KH, Lottenberg R (1994) Characterization of a novel streptokinase produced by Streptococcus equisimilis of non-human origin. Thromb Haemost 72:595-603

18. Ohkuni H, Todome Y, Suzuki H, Mizuse M, Kotani N, Horiu- chi K, Shikama N, Tsugita A, Johnston KH (1992) lmmuno- chemical studies and complete amino acid sequence of the strep- tokinase from Streptococcus pyogenes (group A) M type 12 strain A374. Infect lmmun 60:278-283

19. QIAGEN Plasmid Handbook (1992) QIAGEN GmbH, Hilden, Germany

20. QIAEX Handbook (1993) QIAGEN GmbH, Hilden, Germany 21. Russel RRB, Aduse-Opuku J, Sutcliffe IC, Tao L, Ferretti JJ

(1992) A binding protein-dependent transport system in Strep- tococcus mutans responsible for multiple sugar metabolism. J Biol Chem 267:4631-4637

22. Sambrook J, Fritsch EF, Maniatis TE (eds) (1989) Molecular cloning: a laboratory manual, 2nd edn. Cold Spring Harbor La- boratory, Cold Spring Harbor

23. Steiner K, Malke H (1995) Transcription termination of the streptokinase gene of Streptococcus equisimilis H46A: bidirec- tionality and efficiency in homologous and heterologous hosts. Mol Gen Genet 246:374-380

24. Walter F, Siegel M, Malke H (1989) Nucleotide sequence of the streptokinase gene from a Streptococcuspyogenes type 1 strain. Nucleic Acids Res 17:1261

25. Walter F, Siegel M, Malke H (1989) Nucleotide sequence of the streptokinase gene from a group-G Streptococcus. Nucleic Ac- id Res 17:1262