Experimental Hematology 34 (2006) 1553–1562

Cloning and characterizing mutated human stromal cell-derivedfactor-1 (SDF-1): C-terminal a-helix of SDF-1a plays a critical role

in CXCR4 activation and signaling, but not in CXCR4 binding affinity

Yi Tana,b,d,e, Jun Duc, Shaoxi Caib, Xiaokun Lid,e, Weifeng Mab,Zhigang Guob, Hongyuan Chenb, Zhifeng Huange, Jian Xiaoa,e, Lu Caie,f, and Shaohui Caia

aDepartment of Clinical Pharmacology, Pharmacy School of Jinan University, Guangzhou, China; bCollege of Bioengineering, Key

Laboratory for Biomechanics and Tissue Engineering of the State Ministry of Education, Chongqing University, Chongqing, China;cSchool of Pharmaceutical Sciences, Sun Yat-sen University, Guangzhou, China; dBioengineering Medicine Research Center of the State Ministry

of Education, Jinan University, Guangzhou, China; eSchool of Pharmaceutical Sciences, Wenzhou Medical College, Wenzhou, China; fDepartments of

Medicine, Radiation Oncology and Pharmacology and Toxicology, University of Louisville, Ky., USA

(Received 9 June 2006; revised 28 June 2006; accepted 10 July 2006)

Objective. A novel C-terminal a-helix-defective mutant of human stromal cell-derived factor-1 (SDF-1), hSDF-154, was designed and produced in order to develop an optimal CXC chemo-kine receptor 4 (CXCR4) antagonist.

Materials and Methods. Human native SDF-1 and a-helix defective SDF-1 (hSDF-154) werecloned from human bone marrow stromal cells by reverse transcription polymerase chain reac-tion, inserted into vector pET-30a(+), and transformed into Escherichia coli strain BL21(DE3).The recombinant hSDF-154 was purified and refolded under optimized conditions and itsfunctional characteristics were compared with the native form of SDF-1. Functional evalua-tion includes migration of Jurkat and MOLT4 cells assessed by chemotaxis assay, intracellularcalcium influx in these cells measured by flow cytometry, extracellular signal-regulated kinase(ERK) phosphorylation analyzed by Western blot assay, receptor binding affinity examined bysequential concentrations of unlabeled SDF-1a, hSDF-154 competition with 125I- SDF-1a, andinternalization of CXCR4 on the cell surface detected by flow cytometry.

Results. hSDF-154 had significantly decreased chemotaxic ability, such as cell migration, ascompared to the native hSDF-1. hSDF-154 failed to trigger CXCR4 to induce transient cal-cium influx and ERK phosphorylation. However, both hSDF-154 and the native hSDF-1have similar binding affinity to CXCR4 and a similar ability to induce CXCR4 internalization.

Conclusion. These results indicate that hSDF-154, which has a defective C-terminal a-helix,a normal N-terminus, and a normal central b-strand scaffold structure, retains normal bind-ing affinity to CXCR4 and normal induction of CXCR4 internalization, but fails to activateCXCR4-mediated cellular signaling and chemotaxis. Therefore, the C-terminal a-helix ofhSDF-1 plays a critical role for CXCR4 stimulation. The hSDF-154, which efficiently bindsto and induces internalization of CXCR4 without activating CXCR4-related intracellular sig-naling and cell migration, may serve as an optimal CXCR4 antagonist. � 2006 InternationalSociety for Experimental Hematology. Published by Elsevier Inc.

Stromal cell-derived factor-1a (SDF-1a) is a member of thechemokine family of structurally related proteins with cellchemoattractant activity and is produced constitutivelyand maintained in high levels in bone marrow stromal cells

Offprint requests to: Lu Cai, M.D., Ph.D. or Shaohui Cai, M.D., Ph.D.,

Departments of Medicine, Radiation Oncology and Pharmacology and Tox-

icology, University of Louisville, 511 South Floyd Street, MDR 533, Louis-

ville, KY 40202; E-mail: L0cai001@louisville.edu or csh5689@sina.com

0301-472X/06 $–see front matter. Copyright � 2006 International Society fo

doi: 10.1016/j.exphem.2006.07.001

[1]. SDF-1a has a fundamental role in the trafficking andhoming of bone marrow cells [2,3]. The high level of evolu-tionary conservation of the SDF-1a sequence between spe-cies and the finding that SDF-1a– or CXC chemokinereceptor 4 (CXCR4)-knockout mice have severely deficientmyelopoiesis and lymphopoiesis indicate that SDF-1a playsa critical role in regulating leukocytes, hematopoietic precur-sor migration, pre-B cell proliferation and cerebellar devel-opment [3,4]. In vitro SDF-1a is able to stimulate

r Experimental Hematology. Published by Elsevier Inc.

1554 Y. Tan et al./ Experimental Hematology 34 (2006) 1553–1562

chemotaxis of a wide range of cells, including monocytes andbone marrow hematopoietic progenitor cells (HPCs) [2,5].

SDF-1a is the only known ligand of CXCR4, a seven-transmembrane receptor that has also been named LESTR[6], HUMSTR [7], or fusin [8]. CXCR4 is expressed on he-matopoietic stem and progenitor cells. Through its interac-tion with CXCR4, SDF-1a traffics and retains normalHPCs in bone marrow [6–9]. If SDF-1a levels were increasedin circulation, these HPCs would be peripherally mobilized[10,11]. However, multiple and complicated functions ofthe SDF-1a/CXCR4 axis make it impossible to directly usethe native SDF-1a to peripherally mobilize HPCs. For exam-ple, because SDF-1a traffics not only normal stem cells, butalso tumor stem cells that express CXCR4, the latter wouldincrease the risk of stimulating tumor metastasis [12,13].SDF-1a also interacts with the orphan receptor RDC1 [14]to activate gene expression associated with chondrocytehypertrophy, angiogenesis, and increased matrix degrada-tion, possibly causing early development of osteoarthritis[15]. CXCR4 antagonists are being evaluated for peripheralmobilization of HPCs [16–19]. Although these CXCR4 an-tagonists inhibit the SDF-1a/CXCR4 interaction and donot increase the risk of stimulating tumor metastasis[20,21], a better strategy to develop a new antagonist ofCXCR4 that peripherally mobilizes HPCs, without the riskof tumor metastasis and other pathogens, may be to modifythe original structure of SDF-1a to retain the high affinityfor CXCR4 without activating CXCR4 downstream signaling.

Several studies have indicated that there are three majorstructural regions of the native SDF-1a. The N-terminalresidues and central core regions both play critical rolesin the binding activity of SDF-1a with CXCR4. The C-ter-minal region does not appear to be involved in the bindingactivity of SDF-1a [22–25]. We hypothesized that theC-terminal region may play a critical role in activation ofCXCR4 signaling. A pilot study in which we produceda mutant mouse SDF-1a with deletion of 55 to 67 residuesof C-terminus, named mSDF-1/54R, confirmed our hypoth-esis [26]: mSDF-1/54R had conserved binding affinity forCXCR4, but lost its ability to activate CXCR4 and induceintracellular signaling. However, due to species issues, itwould not be a proper molecule to be further investigatedfor its clinical application. Therefore, the present studywas undertaken with the following objectives: 1) to producea novel mutant of human SDF-1a with deletion of residues55–68 in the C-terminal a-helix, named hSDF-154; 2) tosystemically characterize the biochemical and biophysio-logical features of hSDF-154 in two kinds of cell lines, in-cluding Jurkat cells that were only used in the previous pilotstudy; 3) to determine the competitive binding capacity ofhSDF-154 to CXCR4 with native SDF-1a by direct bindingassay, instead of an indirect assay used in the previously pi-lot study [26], and 4) to compare the binding capacity ofhSDF-154 to CXCR4 with the CXCR4 antibody that wasused in other studies [20].

Materials and methods

Bone marrow tissue, plasmids, E. coli, and cell linesHuman bone marrow was provided by the First Affiliated Hospi-tal, Chongqing Medical University. Expression vector PET30a(þ),cloning E. coli strain JM109 and expression E. coli strainBL21(DE3) were purchased from Novagen. Jurkat cells were giftsfrom Nagoya University Graduate School and Faculty of Medi-cine, Nagoya, Japan, and MOLT4 cells were from the China Cen-ter for Type Culture Collection.

ReagentsReagents used in this study include: total RNA extraction regentTRIZOL Reagent, first-strand synthesis system for reverse-tran-scription polymerase chain reaction (RT-PCR), SuperScript (Invi-trogen, Carlsbad, CA, USA), Taq polymerase, PCR fragmentrecovery kit, DNA ligation kit version2, KpnI and EcoRI, agarose,DL2000þDL15000 DNA Marker, Isopropyl b-D-thiogalactoside(IPTG; TAKARA, Dalian, China), E.Z.N.A. Plasmid MiniprepKit (Omega Bio-tech, British Columbia, Canada), Yeast extractand Tryptone, ProBond Resin for purification of 6xHis-Taggedproteins (Invitrogen), protein marker (Fermantas Life Sciences,Hannover, MD, USA), NOVEX Chromagenic Western blottingdetection system (Novex, San Diego, CA, USA), goat anti-hSDF-1 antibody, alkaline phosphotase rabbit anti-goat IgG(HþL)(Zhongshan, Beijing, China), EKMax Enterokinase (Invitrogen),BCA protein assay kit (Pierce Biotechnology, Inc., Rockford,IL, USA), phycoerythrin (PE) mouse anti-human CXCR4 (fusin)monoclonal antibody (Clone:12G5, R&D, Minneapolis, MN,USA), goat anti-mouse IgG-fluorescein isothiocyanate (FITC)(BOSTER, Wuhan, China), and Fluo-3/AM (Biotium, Hayward,CA, USA). Mouse anti-p-extracellular signal-regulated kinase(ERK) (E-4) and rabbit anti-ERK2 antibodies were purchasedfrom Santa Cruz Biotechnology (Santa Cruz, CA, USA).

Designing primers and cloningthe native hSDF-1a and hSDF-154Primers for cloning human native and mutant of SDF-1a were syn-thesized by Shanghai Sangon Biological Engineering Technologyand Service Co., Ltd. (Shanghai, China) according to cDNA se-quences of native SDF-1a. Their sequences are: sense primers of na-tive hSDF-1a and hSDF-154, including Enterokinase recognizingsite encoding sequences, 50- CATGCCATGGGGTACCCATGAC-GACGACGACAAGAAGCCCGTCAGCCTGAGCTAC-30, anti-sense primer of native SDF-1a: 50-GGGAATTCTTACTTGTTTAAAGCTTTCTCC-30, and antisense primer of hSDF-154:50- GGGAATTCTCACTTCGGGTCAATGCACAC-30.

Total mRNA from human bone marrow was extracted withTRIZOL reagent and the first strand of cDNA was generatedfrom 2 mg total RNA using oligo-dT primer and a SuperScript IIReverse Transcriptase (GIBCO BRL, Grand Island, NY, USA).PCR amplification for hSDF-1a and hSDF-154 cDNA was per-formed initially by 94�C denaturation (3 minutes), followed by30 cycles of three PCR steps of 10 seconds each at 94�C, 55�C,and 72�C, and terminated with an extension prolongation for 5minutes at 72�C. After amplification reaction, the PCR productswere fractioned on 1.2% agarose gel.

Construction of expression vectorThe PCR products of hSDF-1a and hSDF-154 were digested byKpnI/EcoRI and inserted into a bacterial expression plasmid

1555Y. Tan et al. / Experimental Hematology 34 (2006) 1553–1562

pET-30a(þ). The recombinants pET-30a(þ)/hSDF-1a and pET-30a(þ)/hSDF-154 were transformed into the competent bacteriaJM109 and positive clones were selected by 50 mg/L of kanamy-cin. The recombinant expression vectors were extracted and puri-fied by E.Z.N.A. Plasmid Miniprep Kit and sequenced by usinga 310 Genetic Analyzer (ABI Global Medical Instrumentation,Inc., Ramsey, MN, USA) with the primer of T7.

Expression and purification ofthe native hSDF-1a and hSDF-154The positive clones of BL21 (DE3) transformed with pET-30a(þ)/hSDF-1a or pET-30a(þ)/hSDF-154 were amplified in Luria-Ber-tani (LB) medium and induced with IPTG. The purification proce-dure of denaturing conditions was based on the instruction of theuser manual of ProBond Resin. The purified hSDF-1a and hSDF-154 were refolded and concentrated. The amount of target proteinswas examined by sodium dodecyl sulfate-polyacrylamide gel elec-trophoresis (SDS-PAGE). Immunoblot analysis was performed ac-cording to the user manual of NOVEX Chromagenic WesternBlotting Immunodetection System. The verified hSDF-1a andhSDF-154 were digested by EKMax Enterokinase to delete6xHis-Tags (approximate 4.78Kd) and separated by reversephase-high pressure liquid chromatography (RP-HPLC). ThenhSDF-1a and hSDF-154 were lyophilized and the concentrationswere determined by BCA Protein Assay Kit.

N-terminal amino acids sequencingPurified hSDF-1a and hSDF-154 were performed on SDS-PAGEelectrophoresis, transferred onto polyvinylidene difluoride mem-branes and stained with Coomassie Blue. N-terminal amino acidswere sequenced by ABI Procise 492 cLC protein sequencer.

Chemotaxis assayMigration of Jurkat and MOLT4 cells was assessed in disposableTranswell trays (Kurabo, Osaka, Japan) with 6-mm–diameterchambers and membrane pore size of 5 mm. Briefly, hSDF-1aand hSDF-154 were diluted in RPMI 1640 containing 1 mg/mLof bovine serum albumin. Seven-hundred microliters each of di-luted samples was added to the lower wells. To the upper wells,200 mL suspension of Jurkat or MOLT4 cells at a concentrationof 1 � 107/mL were added. For hSDF-154 chemotaxis assay, var-ious concentrations of hSDF-154 were preincubated with thesecells at 37�C for 2 hours. In another experiment, the monoclonalantibody 12G5 was pre-incubated with these cells at 50 mg/mLfor 15 minutes at 4�C. Two-hundred microliters preincubated Ju-rkat cells at a concentration of 1 � 107/mL was also added tothe upper wells. After incubation at 37�C, 5% CO2 for 2 hours,cells that migrated to the lower wells were counted. The controlwas the migrating cells without any treatment [26].

Intracellular calcium measurementMOLT4 cells (1 � 107/mL) were loaded with the calcium indica-tor dye Fluo-3/AM at a final concentration of 4 mmol/L for 45minutes in Ca2þ flux assay buffer (Hank’s balanced salt solutioncontaining 20 mM HEPES and 0.2% BSA [pH 7.4]) at room tem-perature in the dark. Cells were then washed three times with thesame buffer and maintained in darkness until use. For each sam-ple, a 40-second baseline monitoring was performed by flowcytometry. Then sample aspiration was briefly paused and hSDF-1a or hSDF-154 diluted in Ca2þ flux assay buffer was quicklyadded. Two concentrations of hSDF-1a and hSDF-154 were

used, i.e., 50 nM and 500 nM. The former was selected becauseof its common use in other studies (500 ng/mL 5 58.64 nmol/L[27]) while the latter was selected because we want to ensure theefficient dose was used for hSDF-154. The Ca2þ response was mea-sured against the change in green fluorescence intensity of the cellsas a function of time. Analysis was performed with the flow cytom-etry with an air-cooled 488-nm argonion laser and CellQuest soft-ware (Becton Dickinson, Mountain View, CA, USA) [27].

Immunoblot analysisExpression of phosphorylated ERK and ERK2 on MOLT4 andJurkat cells was examined by Western blotting assay. Briefly, celllysates were prepared in a lysis buffer containing 2 mg/mLAprotinin, 2 mg/mL Leupeptin, and 1 mM phenylmethylsulfonylfluoride. Total proteins obtained by centrifugation (14000 g for20 minutes at 4 �C) were size-separated by electrophoresis on12% SDS-PAGE gels and electrophoretically transferred toa PVDF membrane. Nonspecific binding was blocked withTBST containing 5% BSA for 2 hours at 25 �C. The membranewas immunoblotted with primary antibodies (1:1000 dilution ofmouse anti-p-ERK (E-4) or rabbit anti-ERK2) overnight at 4�Cand subsequently exposed to a horseradish peroxidase-conjugatedsecondary antibody (1:1000 dilution in TBST) for 1.5 hours at25�C. Immunoreactive bands were visualized using an electroche-miluminescence detection system.

Receptor-binding assayJurkat cells were used for the binding assay. Cells were washedtwice in phosphate-buffered saline (PBS), and cell pellets were re-suspended in binding buffer (50 mM Tris-HCl [pH 7.4], 1 mMCaCl2, 5 mM MgCl2, and 0.1% BSA) at a concentration of 5 �105 cells/mL. For receptor binding or competition assays [22], var-ious concentrations of unlabeled hSDF-1a (0.1–1000 nM) or hSDF-154 (0.01–1000 nM) were liquated in 50 or 100 mL 125I-SDF-1a(30,000–50,000 cpm, approximate 25 pM, prepared by modifiedChloramine-T-method), 400 mL Jurkat or MOLT4 cells was added,and the final volume was adjusted to 600 mL by binding buffer. Thereaction system was incubated at 4�C for 2 hours with shaking andterminated by addition of 1 mL ice-cold PBS. The cell pellets werecollected by centrifugation at 4�C and washed twice with coldPBS. Cell pellet-associated radioactivity was counted usinga gamma counter. Nonspecific binding was determined in the pres-ence of 1 mM unlabeled hSDF-1a.

CXCR4 internalization assayFor analysis of the amounts of CXCR4 on the cell surface, bothMOLT4 and Jurkat cells were plated into each well of a 24-welltissue culture plate and treated with hSDF-1a or hSDF-154 forthe indicated time. Cells were harvested and incubated with mouseanti-human CXCR4 antibody at 37�C for 30 minutes. After twowashes with PBS, cells were stained with goat anti-mouse IgG-FITC under the same conditions. Cells were analyzed using flowcytometry following two washes with PBS. The control was thecells not incubated with mouse anti-human CXCR4 antibody[26]. In addition to the this method using fluorescent-labeled sec-ond antibody, the amounts of CXCR4 was also detected directlyusing fluorescent-labeled first antibody, PE mouse anti-humanCXCR4 (fusin) monoclonal antibody (Clone: 12G5, R&D, Minne-apolis, MN, USA), which was incubated with MOLT4 cells at

1556 Y. Tan et al./ Experimental Hematology 34 (2006) 1553–1562

hSDF-1α=

hSDF-154=

1 10 20 30 40 50 60 68

Figure 1. Design of a-helix-defective mutant of human stromal cell-derived factor-1 (hSDF-154) prepared by deletion of 55–68 amino acids at the C-ter-

minus of human stromal cell-derived factor-1a (hSDF-1a). The amino acid sequence is shown.

37�C for 30 minutes and then after two washes with PBS, the cellswere analyzed using a flow cytometry.

Statistical analysisAll data was expressed as mean 6 standard deviation. Differencesbetween groups were assessed by one-way analysis of variance. Ifthe variances between groups were homogenous (Levene’s test),groups were subjected to the multiple comparison Bonferroni’stest. A p value ! 0.05 was considered statistically significant.

Results

Cloning and construction ofrecombinants of hSDF-1a and hSDF-154Human SDF-1a (hSDF-1a) and hSDF-154 were designedaccording to the sequence of native human SDF-1a(Fig. 1), which was amplified from human bone marrow

stromal cells by RT-PCR and inserted into expressionvector pET-30a(þ). The recombinants were identified byrestriction enzyme digestion and further confirmed byDNA sequencing (data not shown).

The native hSDF-1a and hSDF-154 proteins were ex-pressed in BL21 (DE3) with induction of IPTG at a finalconcentration of 1 mM, and purified by Ni-Resin immobileaffinity chromatography. The eluted proteins were refoldedby gradually changing the buffer to natural conditions.Purity of the recombinant proteins was determined usingSDS-PAGE analysis with Coomassie Blue staining andconfirmed by Western blot with hSDF-1a specific antibody.Results showed that the molecular weight sizes of bothhSDF-1a and hSDF-154 with His-Tags were as expectedand purities of both hSDF-1a and hSDF-154 were approx-imately 90% (Fig. 2A, B). The 6xHis-Tags were deleted by

A B

32M

45.0

66.0

35.0

14.418.4

25.0

1 4 32M 1 4

C D

I

II

I

II

0

0102030405060708090

100110120130140150160

2 4 6 8 10 12 14 16 18 20 22 24 26 28 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28Times (minutes) Times (minutes)

4.15

24.1

7

15.3

7

4.67

Volta

ge (m

V)

0102030405060708090

100110120130140150

Volta

ge (m

V)

Figure 2. Identification of both human stromal cell-derived factor-1a (hSDF-1a)/His-tag and a-helix-defective mutant of human stromal cell-derived factor-

1 (hSDF-154)/His-tag by sodium dodecyl sulfate polymacrylamide gel electrophoresis (SDS-PAGE) gel and Western blot (top panel), as well as separation of

hSDF-1a and hSDF-154 from His-tag digested by EKMax with reverse phase-high pressure liquid chromatography (bottom panel). (A,B) M, protein marker;

lanes 1 and 2, SDS-PAGE and lanes 3 and 4 Western blot of 1, 2 with goat anti-hSDF-1 antibody. (C,D) I: His-tag; II: hSDF-1a or hSDF-154.

1557Y. Tan et al. / Experimental Hematology 34 (2006) 1553–1562

EKMax Enterokinase and separated by RP-HPLC (Fig. 2C,D). The N-terminal amino acid sequences of hSDF-1a andhSDF-154 were confirmed by a protein sequencer (data notshown).

hSDF-154 loses chemotaxisand CXCR4-mediated signalingInduction of the migration of T lymphocytes through inter-action with CXCR4 is an important physiological functionof the native hSDF-1a, but also a major concern for poten-tial tumor metastasis [5,12,13]. Because Jurkat cells, a Tlymphocyte leukemia cell line, constitutively expressCXCR4 on the cell surface [21], their migration by chemo-taxis assay to hSDF-1a and hSDF-154 was determined.Within the dose range of 1 to 50 nmol/L (p ! 0.05),hSDF-1a induced a dose-dependent chemotaxic effect ofJurkat cells; however, migration was significantly decreasedat higher hSDF-1a concentrations (100 and 1000 nmol/L).hSDF-154 did not induce significant chemotaxis (Fig. 3A),suggesting the importance of the a-helix structure for thechemotaxic activity of hSDF-1a. In addition, similar results

05

1015202530354045505560

*

*

***

**

*

Mig

ratio

n of

Jur

kat (

% o

f inp

ut) A

0

0 0.1 1 10 50 250 --

0 1 2.5 5 10 25 50 100 1000

5

10

15

*

*

*

*

hSDF-1ahSDF-154

hSDF-1ahSDF-154

Mig

ratio

n of

MO

LT4

(% o

f inp

ut)

Concentration (nmol/L)

Concentration (nmol/L)

B

Figure 3. Chemotaxic activity of human stromal cell-derived factor-1a

(hSDF-1a) and a-helix-defective mutant of human stromal cell-derived

factor-1 (hSDF-154). Chemotaxic activity of both hSDF-1a and hSDF-

154 on Jurkat cells (A) and MOLT4 (B) was performed with transwell

trays with a membrane pore size of 5 mm, as described in Materials and

Methods. The background was the mean number of migrating cells without

any treatment. The results are shown as mean 6 standard deviation from

three separate experiments with triplicate experiments for each condition.

*p ! 0.05 vs control.

were also observed in MOTL4 cells at selective dose levelsof hSDF-1a and hSDF-154 (Fig. 3B).

Another key function of hSDF-1a is the increase in in-tracellular free calcium induced by the activation ofCXCR4 [5,26]. To further define whether a-helix–defectivehSDF-1a induces increases in intracellular calcium, the ef-fect of hSDF-154 on intracellular calcium concentrationswas measured. A rapid and transient increase in intracellu-lar calcium concentrations in response to the addition ofhSDF-1a at 50 nmol/L was observed. No future increasein intracellular calcium concentrations when hSDF-1awas added at 500 nmol/L was noted as compared to thatat 50 nmol/L (Fig. 4), suggesting that there is a plateauof the response to hSDF-1a at doses O50 nmol/L. Moreimportantly, no response of the intracellular calcium in re-sponse to hSDF-154 either at low or at high concentrationwas observed (Fig. 4).

Extracellular signal-regulated kinase (ERK) signaling isknown a critical signal transduction pathway associatedwith SDF-1a activation of CXCR4 [28,29]. The effects ofhSDF-1a and hSDF-154 on ERK phosphorylation in bothJurkat and MOLT4 cells were examined by Western blot-ting assay (Fig. 5). Neither hSDF-1a nor hSDF-154 signif-icantly changed total ERK2 expression. hSDF-1a induceda dose-dependent increase in the phosphorylation of ERKin both Jurkat and MOLT4 cells; however, hSDF-154 wasunable to induce ERK phosphorylation (Fig. 5). These find-ings, combined with the migration data, suggest that the a-helix structure of hSDF-1a plays a critical role in activatingCXCR4 and triggering G-protein–coupled signaling cas-cades during cell migration.

Figure 4. a-Helix-defective mutant of human stromal cell-derived factor-

1 (hSDF-154) induced distinct intracellular Ca2þ patterns from that of

human stromal cell-derived factor-1a (hSDF-1a). The effects of both

hSDF-154 and hSDF-1a at indicated concentration on intracellular Ca2þ

influx in MOLT4 cells were measured by flow cytometry analysis with

Fluo-3/AM staining, as described in Materials and Methods. Data are typ-

ical representatives of duplications for each group.

1558 Y. Tan et al./ Experimental Hematology 34 (2006) 1553–1562

hSDF-154 preserves capacity forbinding to and internalizing CXCR4To explore whether hSDF-154 preserves its binding abilityto CXCR4, two experiments were performed. First, whetherhSDF-154 can be used as an antagonist for CXCR4 by in-ducing CXCR4 internalization was investigated usingMOLT4 and Jurkat cells. Both cells were treated withhSDF-1a or hSDF-154, under the conditions indicated inFigure 6A, and the amount of CXCR4 on the cell surfacewas measured by flow cytometry. Results showed thathSDF-154 could induce a pattern of CXCR4 internalizationsimilar to hSDF-1a in MOTL4 cells (Fig. 6A). Similar re-sults were also observed in Jurkat cells (data not shown). Inaddition, the amounts of CXCR4 were also detected di-rectly by fluorescent-labeled anti-human CXCR4 monoclo-nal antibody in MOTL4 cells (Fig. 6B). Dose-dependentinternalization of CXCR4 after treatment with hSDF-154was observed. The effect of hSDF-154 at 500 nmol/L issimilar to that of hSDF-1a at same dose (Fig. 6B).

The second experiment directly examined, by saturableSDF-1a binding assay, the binding affinity of SDF-154with CXCR4 because CXCR4 internalization is dependenton its ligand binding [30]. Binding of 125I-labeled SDF-1ato CXCR4 positive Jurkat cells was competed with eithercold hSDF-1a or hSDF-154. Analysis of the competitioncurves and Scatchard values showed that SDF-1a andSDF-154 have similar IC50 and Kd values: SDF-1a, IC50

5 23.126 67.623 nM, Kd 5 18.496 6 4.761 nM(Fig. 7A) and SDF-154, IC50 5 23.352 6 4.048 nM, Kd

5 20.835 6 2.594 nM (Fig. 7B). The results confirmedthat the defective C-terminal a-helix area did not affectthe binding affinity of SDF-154 to CXCR4, which is consis-tent with the receptor internalization assays. Therefore, theC-terminal a-helix of hSDF-1a is not predominantlyinvolved in CXCR4 binding and CXCR4 internalization.

Control

hSDF-1 hSDF-154

20 10 5 2.5 50 20 (nmol/L)510

p-ERK

ERK2

p-ERK

ERK2

Jurkat

MOLT4

Figure 5. Failure of a-helix-defective mutant of human stromal cell-de-

rived factor-1 (hSDF-154) to trigger extracellular signal-regulated kinase

(ERK) signaling. Both MOLT4 and Jurkat cells were stimulated with or

without human stromal cell-derived factor-1a (hSDF-1a) and hSDF-154

for 2 minutes. Cells were subjected to immunoblotting analysis with

mouse anti-phospho-ERK antibody (labeled as p-ERK) and with rabbit

anti-ERK2 antibody (labeled as ERK2). Images are the representative

gels of at least three experiments for each cell line.

hSDF-154 inhibits hSDF-1a-induced cell migrationWe have confirmed that hSDF-154 preserves its binding toand induces internalization of CXCR4, but is unable to trig-ger cell migration and intracellular calcium influx. Whetherthe hSDF-154 inhibition of CXCR4 activation really pre-vents subsequent reactions of native hSDF-1a withCXCR4 is a critical question for the potential antagonisticapplication of hSDF-154. Therefore, whether hSDF-154has an inhibitory effect on native hSDF-1a–mediated che-motaxis was examined. Both Jurkat and MOLT4 cellswere preincubated with various concentrations of hSDF-154 for 2 hours and then were subjected to chemotaxis as-say with 10 nmol/L hSDF-1a. Migration of cells inducedby hSDF-1a was blocked in a dose-dependent manner bypreincubation of cells with hSDF-154 in Jurkat cells(Fig. 8A) and MOLT4 cells (Fig. 8B). At concentrationshigher than 10 nmol/L of hSDF-154, the native hSDF-1a–mediated cell migration was significantly inhibitedcompared to the cells treated with hSDF-1a alone (p !0.05). Furthermore, the inhibitory effect of hSDF-154(500 nM) on migration is similar to that of anti-CXCR4monoclonal antibody (CXCR4 mAB 12G5 at a concentra-tion of 50 mg/mL [5 8000 nM], Fig. 8C), suggesting abouta 16-fold (8000/500) inhibitory effect of hSDF-154 as com-pared to anti-CXCR4 monoclonal antibody. Furthermore,no synergic effect on migration inhibition was found ifboth hSDF-154 and anti-CXCR4 monoclonal antibody12G5 coexisted (Fig. 8C).

DiscussionIn the present study, recombinant hSDF-154, without theC-terminal a-helix of the native hSDF-1, was designed,expressed, and purified from bacteria and refolded underoptimized conditions. To avoid artificial effects, the nativehSDF-1a was produced using the same procedure ashSDF-154. Compared to the native hSDF-1a, hSDF-154loses biological activity in inducing ERK activation, cal-cium influx and chemotaxis, but still possesses highly spe-cific binding affinity to CXCR4 and induces CXCR4internalization. Pretreatment of the cultured cells withhSDF-154 significantly inhibited the chemotaxic activityof hSDF-1a.

The first important finding of the present study was toconfirm the critical role of the C-terminal a-helix ofhSDF-1a in CXCR4 activation and noncritical role inreceptor binding and internalization. hSDF-1a is composedof three distinct structural regions, detected by cry-stallographic and nuclear magnetic resonance technique:an N-terminus, a central core region of three anti-parallelb-sheets and a C-terminal a-helix [22–25,31]. Previousstudies of synthetic peptides derived from SDF-1a N-termi-nus and SDF-1a mutants demonstrated the essential role ofthe SDF-1a N-terminus, especially the first nine residues,as the major site for direct interaction with the receptor

1559Y. Tan et al. / Experimental Hematology 34 (2006) 1553–1562

Figure 6. Capability of a-helix-defective mutant of human stromal cell-derived factor-1 (hSDF-154) and human stromal cell-derived factor-1a (hSDF-1a)

induction CXC chemokine receptor 4 (CXCR4) internalization. MOLT4 cells were treated with hSDF-1a and hSDF-154 for 2 hours at indicated concentra-

tions and CXCR4 on the cell surface was detected indirectly by fluorescein isothiocyanate (FITC)-labeled second antibody (A) or directly by PE-labeled

CXCR4 antibody (B), using flow cytometry. The level of CXCR4 on the cell surface is shown as mean fluorescence density, as described in Materials

and Methods. MOLT4 cells without any treatment were used as a positive control.

and subsequent signal transduction [22,25,31]. The clusterof positively charged residues on the central b-sheet regionis important for SDF-1a to bind to CXCR4, but not strictlyrequired for the interaction to take place.

In contrast, the functional role of the C-terminus ofSDF-1a is less defined. A few studies seem to reveal thatthe C-terminus plays an enhancing role, rather than a re-quired one, in the binding action of SDF-1a to CXCR4

1560 Y. Tan et al./ Experimental Hematology 34 (2006) 1553–1562

[22–24]. The C-terminal peptide (55–67) alone was foundto neither bind nor activate CXCR4; however, the attach-ment of the SDF-1a C-terminal sequence (residues 55–67) to the N-terminal sequence (residues 5–14) was ableto enhance the induction of calcium influx of the N-termi-nal peptide, suggesting the importance of the C-terminal se-quence, particularly residues of 55-67, in SDF-1a/CXCR4downstream signaling effect [22–24]. To further elucidatethe functional effects of the C-terminal a-helix of SDF-1,hSDF-1a with deletion of residues of 55–68, hSDF-154,showed a complete loss of SDF-1a/CXCR4 downstreamsignaling effectsdsuch as ERK activation, calcium influx,and cell migrationdwith preservation of their native bind-ing affinity to CXCR4. The latter was shown by direct re-ceptor binding assay and its competing prevention of

0

-10

-11 -10

-9 -8 -7 -6 -5

20

40

60

80

100

120

Boun

d of

125 I

-SD

F-1α

% o

f con

trol

Boun

d of

125 I

-SD

F-1α

% o

f con

trol

Competitor hSDF-1α (LogM)

Competitor hSDF-154 (LogM)

IC50=23.126 ± 7.623 nM

0.0 0.2 0.4 0.6 0.8 1.0

0.0 0.2 0.4 0.6 0.8 1.0 1.2

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08Kd=18.496 ± 3.611 nM

Boun

d/Fr

eeBound (nM)

A

-9 -8 -7 -6 -5

0

20

40

60

80

100

IC50=23.352 ± 4.048 nM

0.02

0.03

0.04

0.05

0.06

0.07

0.08 Kd=20.835 ± 2.594nM

Boun

d/Fr

ee

Bound (nM)

B

Figure 7. Binding characteristics for human stromal cell-derived factor-

1a (hSDF-1a) and a-helix-defective mutant of human stromal cell-derived

factor-1 (hSDF-154). Displacement curves show inhibition of 125I-SDF-1a

by increasing concentrations of unlabeled SDF-1a (A) and SDF-154 (B) in

Jurkat cells. Values of inhibition IC50 are shown in the lower side of each

graph. The insert presents data as transformed by Scatchard analysis and

the dissociation constant (Kd) for each is shown in the upper right corner

of each graph. Results are shown as mean 6 standard deviation from dif-

ferent experiments with triplicate samples for each condition.

Jurkat or MOLT4 cell migration induced by the nativehSDF-1a. More importantly, the degree of hSDF-154 com-peting with hSDF-1a–induced cell migration is muchhigher than that of CXCR4 antibody 12G5 (Fig. 8).

hSDF-154 loses the potency to activate CXCR4 but stillretains binding affinity to the receptor and the ability to

B

-9 -8 -7 -60

5

10

15

Mig

ratio

n of

MO

LT4

(% o

f inp

ut)

Concentration of hSDF-154 (LogM)

* *, #

*, #

*, #

A

-10 -9 -8 -7 -6 -50

5

10

15

20

*

Pretreatment with hSDF-154 and hSDF-1αTreatment with hSDF-1α aloneNontreatment control

Pretreatment with hSDF-154 and hSDF-1αTreatment with hSDF-1α aloneNontreatment control

Mig

ratio

n of

Jur

kat

(% o

f inp

ut)

Concentration of hSDF-154 (LogM)

*

*, #

*, ## #

Mig

ratio

n of

Jur

kat

(% o

f inp

ut)

Control

hSDF-1a

hSDF-1a+12G5

hSDF-154

hSDF-154+12G5

hSDF-1a+hSDF-15405

101520253035 *

*, #*, #

*, # *, #

C

Figure 8. Completive inhibition of human stromal cell-derived factor-1a

(hSDF-1a)-induced cell migration by a-helix-defective mutant of human

stromal cell-derived factor-1 (hSDF-154) and anti- CXC chemokine recep-

tor 4 (CXCR4) antibody. Both Jurkat (A) and MOLT4 (B) cells were used

to examine the inhibitory effect of hSDF-154 on hSDF-1a–induced cell

migration. In addition, the cell migration activated by hSDF-1a (10 nM)

with pretreatments of either hSDF-154 (500 nM), CXCR4 monoclonal an-

tibody 12G5 (50 mg/mL) alone or combined both was examined using Ju-

rkat cells (C). Three separate experiments with triplicate same for each

condition were performed to collect data, which were shown as the

mean 6 standard deviation. *p ! 0.05 vs control; #p ! 0.05 vs hSDF-

1a alone (positive control).

1561Y. Tan et al. / Experimental Hematology 34 (2006) 1553–1562

induce receptor internalization. This is because hSDF-154still contains both N-terminus and b-strands that are en-gaged in receptor binding and molecule scaffold, respec-tively. The N-terminus is a prerequisite, but not sufficient,for receptor binding and activation, requiring the three an-tiparallel b-strands to facilitate the interaction betweenSDF-1a and CXCR4. The packing requirements of the a-helix in SDF-1a are fulfilled by Trp57, Tyr61, and Leu62,which interact with residues of the first and second b-strands [22]. Perhaps because these interactions betweenthe C-terminal a-helix and b-strands are critical for main-taining the active conformation of SDF-1a to trigger signaltransduction, but not necessary for SDF-1a binding withCXCR4, hSDF-154 reserves its high affinity to the receptor,but fails to activate receptor reaction.

CXCR4 downstream signal transduction is independentof receptor endocytosis [32,33]. For example, the third in-tracellular loop (ICL3) of CXCR4 is specifically involvedin G(i)-dependent signals, such as ERK activation andcalcium mobilization, but does not trigger CXCR4 inter-nalization after SDF-1 binding, indicating that ERKphosphorylation is independent of CXCR4 endocytosis[32]. Therefore, although hSDF-154 with the core scaffoldof SDF-1 reserves its high affinity to the receptor and in-duces internalization, it does not activate CXCR4 to triggerERK phosphorylation, calcium influx, and cell migration.

A second important finding of this study is its potentialapplication for the clinic. Because SDF-1a traffics andhomes normal HPCs in bone marrow through SDF-1a/CXCR4 interaction, inhibition of the SDF-1a/CXCR4interaction, using CXCR4 antagonists to mobilize bonemarrow HPCs into circulation, are being evaluated [16–19].The native SDF-1a traffics HPCs and tumor stem cells thatexpress CXCR4; therefore, blocking the SDF-1a/CXCR4interaction, using monoclonal antibody or antagonists ofCXCR4, may be an alternative approach to prevent metastasisof CXCR4-positive cancer cells in several human tumors[12,13,21]. In addition, because CXCR4 is a main co-receptorby T-cell–tropic human immunodeficiency virus-1 (HIV-1)strains for entry into their target cells, CXCR4 antibodiesand antagonists, including peptides and other small mole-cules, have been explored for potential use to block HIV-1infection [22,34–36]. Among these antagonists, AMD3100has been proven to be a highly specific CXCR4 antagonist,which consistently blocks CXCR4-tropic (X4) HIV-1 viralreplication in all target cell types evaluated so far. However,several studies have shown its toxicity and consequently, ithas no longer been thought of as a relevant anti-HIV agent[34,35]. However, new CXCR4 antagonists are being devel-oped, such as AMD070 and KRH 2731 [36]. Therefore,whether hSDF-154 can be used to peripherally mobilizeHPCs is an interesting issue.

Regarding clinical use of chemokines, antibodies, pep-tides, and small molecules, there will always be several is-sues to be concerned with, including immune response and

side effects [26,31,34,35]. Therefore, the recombinanthSDF-154 may be an alternative candidate to be used forthe peripheral mobilization of HPCs, and the preventionof tumor metastasis or HIV infection. Compared to the na-tive chemokines, antibodies, peptides, and small molecules,hSDF-154 has an essential feature: hSDF-154 is directlyderived from human native SDF-1a and contains bothN-terminus and three b-sheet molecular scaffolds whichmake it possible to keep hSDF-154 in the native Greekkey-like arrangement [22,37]. This distinct tertiary foldand structure may be the main mechanism responsible forthe both binding affinity to and internalization of theCXCR4, as discussed above.

AcknowledgmentsWe thank Dr. Min Zhang (Sichuan University) for his ex-cellent technical assistance in receptor binding assay, andDr. Jian Wang (Chongqing University) for protein samplepreparation. This project was supported, in part, by grantsfrom National Natural Science Foundation of China (No30271519 and 30572209), Visiting Scholar Foundation ofKey Lab of the State Ministry of Education in ChongqingUniversity and Wenzhou Medical College CollaboratingFund (5010 Project).

References1. Baggiolini M, Dewald B, Moser B. Human chemokines: an update.

Annu Rev Immunol. 1997;15:675–705.

2. Aiuti A, Webb IJ, Bleul C, Springer T, Gutierrez-Ramos JC. The che-

mokine SDF-1 is a chemoattractant for human CD34þ hematopoietic

progenitor cells and provides a new mechanism to explain the mobili-

zation of CD34þ progenitors to peripheral blood. J Exp Med. 1997;

185:111–120.

3. Nagasawa T, Hirota S, Tachibana K, et al. Defects of B-cell lympho-

poiesis and bone-marrow myelopoiesis in mice lacking the CXC che-

mokine PBSF/SDF-1. Nature. 1996;382:635–638.

4. Tachibana K, Hirota S, Iizasa H, et al. The chemokine receptor

CXCR4 is essential for vascularization of the gastrointestinal tract.

Nature. 1998;393:591–594.

5. Bleul CC, Fuhlbrigge RC, Casasnovas JM, Aiuti A, Springer TA. A

highly efficacious lymphocyte chemoattractant, stromal cell-derived

factor 1 (SDF-1). J Exp Med. 1996;184:1101–1109.

6. Loetscher M, Geiser T, O’Reilly T, Zwahlen R, Baggiolini M, Moser B.

Cloning of a human seven-transmembrane domain receptor, LESTR,

that is highly expressed in leukocytes. J Biol Chem. 1994;269:232–237.

7. Federsppiel B, Melhado IG, Duncan AM, et al. Molecular cloning of

the cDNA and chromosomal localization of the gene for a putative

seven-transmembrane segment (7-TMS) receptor isolated from human

spleen. Genomics. 1993;16:707–712.

8. Feng Y, Broder CC, Kennedy PE, Berger EA. HIV-1 entry cofactor:

functional cDNA cloning of a seven-transmembrane, G protein-cou-

pled receptor. Science. 1996;272:872–877.

9. Aiuti A, Tavian M, Cipponi A, et al. Expression of CXCR4, the recep-

tor for stromal cell-derived factor-1 on fetal and adult human lympho-

hematopoietic progenitors. Eur J Immunol. 1999;29:1823–1831.

10. Moore MA, Hattori K, Heissig B, et al. Mobilization of endothelial

and hematopoietic stem and progenitor cells by adenovector-mediated

elevation of serum levels of SDF-1, VEGF, and angiopoietin-1. Ann N

Y Acad Sci. 2001;938:36–45.

1562 Y. Tan et al./ Experimental Hematology 34 (2006) 1553–1562

11. Perez LE, Alpdogan O, Shieh JH, et al. Increased plasma levels of

stromal-derived factor-1 (SDF-1/CXCL12) enhance human thrombo-

poiesis and mobilize human colony-forming cells (CFC) in NOD/S-

CID mice. Exp Hematol. 2004;32:300–307.

12. Muller A, Homey B, Soto H, et al. A. Involvement of chemokine

receptors in breast cancer metastasis. Nature. 2001;410:50–56.

13. Kucia M, Reca R, Miekus K, et al. Trafficking of normal stem cells

and metastasis of cancer stem cells involve similar mechanisms: piv-

otal role of the SDF-1-CXCR4 axis. Stem Cells. 2005;23:879–894.

14. Balabanian K, Lagane B, Infantino S, et al. The chemokine SDF-

1/CXCL12 binds to and signals through the orphan receptor RDC1

in T lymphocytes. J Biol Chem. 2005;280:35760–35766.

15. Jones SW, Brockbank SM, Mobbs ML, et al. The orphan G-protein

coupled receptor RDC1: evidence for a role in chondrocyte hypertro-

phy and articular cartilage matrix turnover. Osteoarthritis Cartilage.

2006;14:597–608.

16. Liles WC, Broxmeyer HE, Rodger E, et al. Mobilization of hemato-

poietic progenitor cells in healthy volunteers by AMD3100,

a CXCR4 antagonist. Blood. 2003;102:2728–2730.

17. Zhong R, Law P, Wong D, Merzouk A, Salari H, Ball ED. Small pep-

tide analogs to stromal derived factor-1 enhance chemotactic migra-

tion of human and mouse hematopoietic cells. Exp Hematol. 2004;

32:470–475.

18. Broxmeyer HE, Orschell CM, Clapp DW, et al. Rapid mobilization

of murine and human hematopoietic stem and progenitor cells

with AMD3100, a CXCR4 antagonist. J Exp Med. 2005;201:1307–1318.

19. Pelus LM, Bian H, Fukuda S, Wong D, Merzouk A, Salari H. The

CXCR4 agonist peptide, CTCE-0021, rapidly mobilizes polymorpho-

nuclear neutrophils and hematopoietic progenitor cells into peripheral

blood and synergizes with granulocyte colony-stimulating factor. Exp

Hematol. 2005;33:295–307.

20. Sun YX, Schneider A, Jung Y, et al. Skeletal localization and neutral-

ization of the SDF-1(CXCL12)/CXCR4 axis blocks prostate cancer

metastasis and growth in osseous sites in vivo. J Bone Miner Res.

2005;20:318–329.

21. Tamamura H, Hori A, Kanzaki N, et al. T140 analogs as CXCR4 an-

tagonists identified as anti-metastatic agents in the treatment of breast

cancer. FEBS Lett. 2003;550:79–83.

22. Crump MP, Gong JH, Loetscher P, et al. Solution structure and basis

for functional activity of stromal cell-derived factor-1; dissociation of

CXCR4 activation from binding and inhibition of HIV-1. EMBO J.

1997;16:6996–7007.

23. Luo J, Luo Z, Zhou N, Hall JW, Huang Z. Attachment of C-terminus

of SDF-1 enhances the biological activity of its N-terminal peptide.

Biochem Biophys Res Commun. 1999;264:42–47.

24. Tudan C, Willick GE, Chahal S, et al. C-terminal cyclization of an SDF-

1 small peptide analogue dramatically increases receptor affinity and

activation of the CXCR4 receptor. J Med Chem. 2002;45:2024–2031.

25. Gozansky EK, Louis JM, Caffrey M, Clore GM. Mapping the binding

of the N-terminal extracellular tail of the CXCR4 receptor to stromal

cell-derived factor-1alpha. J Mol Biol. 2005;345:651–658.

26. Cai SH, Tan Y, Ren XD, Li XH, Cai SX, Du J. Loss of C-terminal al-

pha-helix decreased SDF-1alpha-mediated signaling and chemotaxis

without influencing CXCR4 internalization. Acta Pharmacol Sin.

2004;25:152–160.

27. Princen K, Hatse S, Vermeire K, De Clercq E, Schols D. Evaluation of

SDF-1/CXCR4-induced Ca2þ signaling by fluorometric imaging plate

reader (FLIPR) and flow cytometry. Cytometry A. 2003;51:35–45.

28. Bajetto A, Barbero S, Bonavia R, et al. Stromal cell-derived factor-

1alpha induces astrocyte proliferation through the activation of extra-

cellular signal-regulated kinases 1/2 pathway. J Neurochem. 2001;77:

1226–1236.

29. Zhao M, Discipio RG, Wimmer AG, Schraufstatter IU. Regulation of

CXCR4-mediated nuclear translocation of extracellular signal-related

kinases 1 and 2. Mol Pharmacol. 2006;69:66–75.

30. Hesselgesser J, Liang M, Hoxie J, et al. Identification and character-

ization of the CXCR4 chemokine receptor in human T cell lines:

ligand binding, biological activity, and HIV-1 infectivity. J Immunol.

1998;160:877–883.

31. Heveker N, Montes M, Germeroth L, et al. Dissociation of the signal-

ling and antiviral properties of SDF-1-derived small peptides. Curr

Biol. 1998;8:369–376.

32. Roland J, Murphy BJ, Ahr B, et al. Role of the intracellular domains of

CXCR4 in SDF-1-mediated signaling. Blood. 2003;101:399–406.

33. Haribabu B, Richardson RM, Fisher I, et al. Regulation of human che-

mokine receptors CXCR4. Role of phosphorylation in desensitization

and internalization. J Biol Chem. 1997;272:28726–28731.

34. Hendrix CW, Collier AC, Lederman MM, et al. Safety, pharmacoki-

netics, and antiviral activity of AMD3100, a selective CXCR4 recep-

tor inhibitor, in HIV-1 infection. J Acquir Immune Defic Syndr. 2004;

37:1253–1262.

35. Este JA. Virus entry as a target for anti-HIV intervention. Curr Med

Chem. 2003;10:1617–1632.

36. Castagna A, Biswas P, Beretta A, Lazzarin A. The appealing story of

HIV entry inhibitors: from discovery of biological mechanisms to drug

development. Drugs. 2005;65:879–904.

37. Dealwis C, Fernandez EJ, Thompson DA, Simon RJ, Siani MA, Lolis

E. Crystal structure of chemically synthesized [N33A] stromal cell-de-

rived factor 1alpha, a potent ligand for the HIV-1 ‘‘fusin’’ coreceptor.

Proc Natl Acad Sci U S A. 1998;95:6941–6946.