PROTEIN

EXPRESSION

AND PURIFICATION

2, 15-23 (1991)

High-Yield Purification of HIV-1 Proteinase Expressed by a Synthetic Gene in Escherichia C O /I’ Laura Goobar,*,l U. Helena Danielson,? Peter Brodin,$ Thomas Grundstrijm,$ Bo 6berg,* and Erling Norrby* *Department of Virology, Karolinska Institute, S-105 21 Stockholm, Sweden; TDepartment of Biochemistry, Biomedical Center, University of Uppsala, S-751 23, Uppsala, Sweden; and SDepartment of Applied Cell and Molecular Biology, University of Urned, S-901 87 Urned, Sweden

Received June 6, 1990, and in revised form November

5,199O

tase, ribonuclease H, and endonuclease activity (5). The 55kDa Gag precursor polyprotein carries the sequences for the matrix (p17), the capsid (p24), and two nucleocapsid proteins p7 and p6 (6,7). The HIV-l proteinase catalyzes its own cleavage from the Gag-PO1 precursor (8,9) and is responsible for the processing of Gag and Gag-PO1 polyproteins into mature proteins (10-14). The proteinase has been shown to be essential for viral maturation and infectivity (1516) and has therefore been proposed as a potential target for anti-HIV therapy. The purification (17-24) and crystallization (25,26) of a synthetic and a bacterially expressed HIV-l proteinase have been described. This has confirmed its classification as an aspartic type proteinase of homodimeric structure (27,28) each monomer having a calculated molecular mass of 10,774 (29). We have developed a new purification method for bacterially expressed HIV-l proteinase that gives higher yields of proteinase per milligram bacterial protein than the methods previously reported. This purification 0 1991 Academic Press, Inc. method contains only two chromatographic steps and is devoid of time consuming steps such as repeated dialysis. The identity of the purified HIV-l proteinase has The human immunodeficiency virus type 1 (HIV-l)’ been assessedby immunoblot analysis with anti-peptide structural proteins and enzymes are translated as two antibodies and by N-terminal amino acid sequence delarge precursor polyproteins called Gag and Gag-PO1 termination. Its proteolytic activity has been measured (1,2). The Gag-PO1 polyprotein contains, apart from a by cleavage of a peptide substrate containing a natural truncated Gag, a proteinase (3,4), a reverse transcripcleavage site for the HIV-l proteinase and by cleavage of denatured bovine serum albumin.

A rapid and simple purification procedure for human immunodeficiency virus type 1 (HIV-l) proteinase from a synthetic gene expressed in Escherichia coli has been developed. The synthetic gene was constructed from oligonucleotides containing several restriction enzyme sites in order to allow simple construction of homologous genes. The protein was translated as a precursor which was autocatalytically processed into the mature protein as shown by N-terminal sequence analysis of the purified protein. Immunoblot analysis was used to verify the nature of the expression product and it was found that 2 of 10 anti-peptide antibodies, covering the whole proteinase sequence, were able to react with the enzyme in crude bacterial lysates. These two anti-peptide antibodies represent a continuous sequence partially overlapping the active site. The purification involves two initial precipitation steps followed by cation-exchange and size-exclusion chromatography. A high yield and a high specific activity were achieved.

i To whom correspondence should be addressed at Department of Virology, Karolinska Institute, % SBL Lundagatan 2, S-105 21 Stockholm, Sweden. ’ Abbreviations used: AMV, avian myeloblastosis virus; BSA, bovine serum albumin; DTT, dithiothreitol; HIV-l, human immunodeficiency virus type 1; KLH, keyhole limpet hemocyanin; Mes, 2(Nmorpholino)ethanesulfonic acid; NaOAc, sodium acetate; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; PMSF, phenylmethylsulfonyl fluoride; SDS, sodium dodecyl sulfate. 1046-5928191 $3.00 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

MATERIALS

AND

METHODS

Synthesis and Expression

of HIV-l

Proteinase

Gene

O ligonucleotides were synthesized and assembled into linearized bacteriophage Ml3 vectors using the shotgun ligation technique (30). The complete gene was 15

16

GOOBAR ET AL.

cloned into the runaway plasmid pRCB1 and expressed in Escherichia coli MM294 as previously described for a calbindin D,, gene (31). The cells were harvested by centrifugation, resuspended in 20 m M Tris-HCl, pH 6.5, 1 m M PMSF, 1 m M EDTA, 1 m M DTT to a protein concentration of 4-7 mg/ml, and lysed in a French press. Bacterial lysates were kept at -70°C. Production

of Antibodies

A series of 10 peptides was designed to cover the entire sequence of HIV-l proteinase, each peptide being 15 amino acids in length and overlapping each of the neighboring peptides by 5 amino acids. A carboxy-terminal cysteine was added to each peptide to allow coupling to keyhole limpet hemocyanin (KLH) (32). Peptides were obtained from Richard A. Houghten, La Jolla, California (33). Rabbits were immunized with 140 pg of KLHcoupled peptides dissolved in 3.2 ml of equal volumes of phosphate-buffered saline (PBS) and complete Freund’s adjuvant. Two boosts were given at Days 14 and 21 with the same amount of peptide in equal volumes of PBS and incomplete Freund’s adjuvant (3.2 ml total volume). Rabbits were bled at 38, 46, 53, and 60 days and sera were stored at -70°C until use. Electrophoresis

and Immunoblotting

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed essentially as previously described (34). Loading buffer (0.7 m M bromphenol blue, 16% (w/v) SDS, 40% (v/v) 2-mercaptoethanol, and 60 m M Tris-HCl, pH 6.8) in a proportion of 1~1 per 5-~1 sample was added to samples before heat denaturation. Samples (15 ~1) were electrophoresed on a discontinuous SDS polyacrylamide gel (0.75 X 60 X 80 mm) in a Midget electrophoresis unit (Pharmacia LKB Biotechnology, Uppsala, Sweden). Gels were stained with Coomassie brilliant blue (34) or silver (35). Immunoblots were performed according to the method of Small et al. (36), with minor modifications. Proteins from SDS gels were transferred to Immobilon membranes (Millipore Intertech, Bedford, MA) in a Midget multiblot electrophoretic transfer unit (Pharmacia LKB Biotechnology). Protein transfer was carried out for 45 min at 150 mA constant current in transfer buffer (0.7 M glycine, 25 m M Tris, pH 7.7). Membranes were blocked overnight in 5% nonfat dry milk, 0.05% sodium azide in PBS at 4°C. Antisera diluted 1:lOO with PBS were incubated with membranes for 2 h at 37°C rinsed four times with PBS, and incubated for 2 h at 37°C with horseradish peroxidase-conjugated swine immunoglobulin to rabbit immunoglobulin (Dakopatts AB, Denmark). Membranes were rinsed three times with PBS at room temperature and developed by incubation with 4-chloro-1-naphthol (0.5 mg/

ml, 15 methanol/PBS, temperature.

0.01% H,O,) for 30 min at room

Protein Purification The bacterial lysate corresponding to 30 g of cells was thawed on ice and centrifuged at 4Y! for 20 min at 10,OOOg. The supernatant was further centrifuged at 4°C for 30 min at 100,OOOg and stored as aliquots at -70°C or immediately precipitated with acetone. Four milliliters of supernatant corresponding to 1.2 g of cells was precipitated in 9 vol of acetone at -70°C under continuous stirring, centrifuged at 2°C for 15 min at lO,OOOg, and dried under reduced pressure. The dried precipitate could be stored at -7O”C, or immediately used for further purification. Acetone precipitate from 4 ml of supernatant was directly mixed with 3 ml sodium acetate buffer (50 m M NaOAc, 1 m M DTT, 1 m M EDTA, 10% glycerol, pH 5.5), containing 1.2 M ammonium sulfate on ice under continuous stirring for 1 h, and the precipitate was recovered by centrifugation at 4°C for 15 min at 10,OOOg.The pellet was dissolved in 2 ml Mes buffer (50 m M 2-(N-morpholino)ethanesulfonic acid (Mes), 1 m M DTT, 1 m M EDTA, 10% glycerol, pH 6.5) and centrifuged at 4°C for 15 min at lO,OOOg, and 1.5 ml of the supernatant was applied to a CM-Sepharose Fast Flow column (1.6 cm 4 X 8 cm, Pharmacia LKB Biotechnology) equilibrated in the same buffer. The proteinase was eluted from the column at a flow rate of 1 ml/min by increasing the NaCl concentration from 0 to 1 M in 30 ml. The fractions containing proteinase activity (6 ml) were concentrated to 600-800 ~1 by ultrafiltration in an Amicon filter (Amicon B.V. Oosterhout, Holland) with a molecular weight cutoff of 5000. The material (500 ~1) was applied to a size-exclusion chromatography column (Superose 12, HR 10130, Pharmacia LKB Biotechnology) equilibrated with Mes buffer containing 150 m M NaCl. Elution was performed at a flow rate of 0.5 ml/ min with the same buffer. The fractions containing proteinase activity were stored as aliquots at -70°C. For “large-scale purification,” 48 ml supernatant from bacterial lysate corresponding to 14 g cells was precipitated in acetone and ammonium sulfate as described above and resuspended in 24 ml Mes buffer. The sample was then applied to a CM-Sepharose column (2.6 cm 4 X 40 cm) and eluted at a flow rate of 3 ml/min with a NaCl gradient of 0 to 1 M in 300 ml. Pooled fractions containing proteinase activity were concentrated to 1 ml as described above and applied to a Superose 12 column (1.6 cm 4 X 70 cm). Elution was performed at a flow rate of 1.5 ml/min with Mes buffer containing 150 m M NaCl. Protein

Concentration

Protein concentration was determined Rad protein assay dye reagent (Bio-Rad

with the BioLaboratories,

PURIFICATION

Richmond, standard.

CA), using bovine serum albumin

Proteolytic

Activity

OF HIV-l

(BSA) as a

Assays

Proteolytic activity was measured using BSA as substrate according to a method modified from Dittmar and Moelling (37). Samples of 10 ~1 proteinase were added to 20 ~1 heat-denatured BSA (2 mg/ml in H,O) and 20 ~1 of 100 mM NaOAc, pH 5.5,375 mM NaCl, 2 mM EDTA, 2 mM DTT, 20% glycerol. Incubations were routinely carried out overnight (16 h) at 37°C. Reactions were stopped by addition of 10 ~1 loading buffer and heat denaturation, samples were analyzed by SDSPAGE as described above. Cleavage of the peptide substrate SQNYPIVQ-amide (obtained from Goran Utter, Karolinska Institute, Stockholm, Sweden (33)) was performed in a similar manner. Samples of 20 ~1 proteinase were added to 40 ~1 peptide (1 mg/ml in H,O) and 40 ~1100 mM NaOAc, pH 5.5,375 mM NaCl, 2 mM EDTA, 2 mM DTT, 20% glycerol. Incubations were carried out for 1 h at 37°C. Reactions were stopped by the addition of 100 ~1 of 0.2% trifluoroacetic acid. Peptide cleavage was analyzed by reversed-phase HPLC using a 4.0 X loo-mm Cl8 superpat cartridge column (Pharmacia LKB Biotechnology). Elution was carried out with a 15-min linear gradient of 0 to 30% of acetonitrile in H,O, 0.1% trifluoroacetic acid. Peptides were detected by absorbance at 214 nm and quantified by integration of peak areas. The specific activity was calculated as the amount of product (nmol) formed per minute per milligram protein. Amino-Terminal

Sequence Determination

The amino acid sequence was determined with a 477 A protein sequencer equipped with an on-line analyzer (120 A) for liberated phenylthiohydantoin amino acids. The instrument was operated according to the manufacturer’s recommendations (Applied Biosystems, Foster City, CA). The proteinase was precipitated in 38% trichloroacetic acid and dissolved in 0.02% trifluoroacetic acid, 0.3% SDS, before applying it to the sequencer. RESULTS

Expression

of HIV-l

Proteinase

Gene

A synthetic gene for HIV-l proteinase (l), optimized for expression in E. coli, was constructed in order to obtain a system for high-yield production of HIV-l proteinase. The gene (Fig. 1) was designed in a modular way from 22 oligonucleotides with several restriction enzyme sites to facilitate the construction of proteinase molecules with changed amino acid sequence. Codons found to be preferentially utilized in highly expressed E. coli genes were used, but a few moderately used E. coli codons were chosen when restriction enzyme sites were

17

PROTEINASE

to be introduced or avoided. The initiator methionine codon is followed by an AAT codon, reported to be the most efficient second codon in an E. coli gene (38). This asparagine codon is followed by a sequence coding for seven amino acids of the HIV-l Gag-PO1 polyprotein before the autocleavage site (8,9) between phenylalanine and the N-terminal proline of the proteinase. A ribosome binding site with high homology to the 3’ end of 16 S ribosomal RNA was placed at the most commonly found distance between this homology and the initiation codon (39). The gene was expressed in E. coli from pRCB1, a runaway vector with temperature-sensitive replication control and the strong inducible tacpromoter. HIV-l proteinase synthesis was detected by SDSPAGE (Fig. 2A). One band of approximately 10,000 Da was present in lysates of bacteria expressing the HIV-l proteinase and absent in control lysates of bacteria with the pRCB1 plasmid that lacked the HIV-l proteinase gene and contained a gene encoding calbindin D,, (hf, 8500). A mutant with the active site aspartate changed to asparagine (Fig. 1) was constructed; this gave a band of slightly higher molecular weight in SDS-PAGE (data not shown). These results suggest that the complete gene is expressed and that the proteinase is autoproteolytically processed to its native size. Identification

of HIV-l

Proteinase

by Immunoblot

Polyclonal rabbit antibodies were raised against 10 peptides (16 amino acids each), with partially overlapping sequences which covered the entire HIV-l proteinase amino acid sequence. Only two peptides, LDTGADDTVLEEMSLC and EEMSLPGRWKPKMIGC, gave rise to antibodies that reacted in immunoblots with a protein band of molecular weight between 9000 and 11,000, which is in agreement with the molecular weight of the denatured proteinase (10,774) (Fig. 2B). A single band was observed, indicating that the enzyme is completely processed to its native size. One of these antibodies was directed against the active site of the proteinase and the immediately adjacent sequence on the carboxyterminal side; the other one was directed against a sequence which extended toward the carboxy-terminal side of the proteinase. A third peptide, KIGGQLKEALLDTGAC, which also contained the active site sequence but extended toward the amino-terminal side, did not give antibodies which reacted specifically with the HIV-l proteinase in immunoblot (data not shown). No reaction with proteins of molecular weight under 40,000 was detected in proteins from control lysates made from bacteria transformed with plasmids lacking the proteinase coding region (Fig. 2B). High-molecular-weight bands were identified as unspecific binding of antibodies to bacterial proteins as they were all present in both control and proteinase-expressing bacterial lysates (Fig. 2B).

18

GOOBAR

met ssn

sly

sls

leu

thr

vsl

ser

ET AL.

pho

rsn

pho

v

pro

gln

i10

thr

1~

a)n flu

1~

sip thr gly sls asp asp thr vsl 1.u olu QlU L 1 c;laA GCT CTG CTG GA1 ACC GGT GCA al GAT ACC GTA CTG GAA GAA -cTT CGAIGAC GAC CTA TGG CCA CGT CTA CTA TGG CAif GAC CTT CTT

‘12

VO rg trp

gly

gly

phe

ile

vrl

lys

rrg

lys

pro

lys

m

ile

Qly

gly

ile

gin

tyr

ssp

gin

ila

IOU ile

glu 119

I17 %T GGT TTT ATC Ad9 CCA CCA AAA TAG 777

pro

val

ssn

ile

ile

GTA CGT CAG TAT GAT CAG ATC CTC ATCIGi& CAT GCA GTCIATA CTA GTC TAG GAG TAG CTT

gly

srg

ssn

IOU IOU thr

gin

ile

gly

cyt

CCT GTC AAC ATC ATT GOT CGT AAC CTG CTG ACT CA6 ATT GGT TGC WA CAG TTG TAG TAA CCA GCA TTG GAC GAC TGA GTC TAA CCA ACt$ thr

IOU ssn

pho

ACT CT0 AAC ITT TA TGA GAC TIC AAA AT

FIG. 1. Amino acid sequence of an HIV-l proteinase isolate (1) and the nucleotide sequence designed for the synthetic gene. The oligonucleotides used for the construction are numbered and restriction enzyme recognition sites are indicated. The autocleavage site (v) of the proteinase and the Asp to Asn active site mutation are indicated.

Purification

of HIV-l

Proteinase

The proteinase was found only in the soluble fraction of the bacterial lysate, after it had been centrifuged twice to obtain a clear supernatant. Two precipitation steps were used to concentrate and partially purify the proteinase. Protein precipitation was observed at acetone concentrations equal to or higher than 55% (v/v), but this acetone concentration was not sufficient to precipitate the proteinase. At 90% (v/v) acetone, however, a complete recovery of the proteinase was achieved. Partial precipitation of the HIV-l proteinase was observed at 0.6 M ammonium sulfate, but 1.2 M ammonium sulfate was required for a complete recovery of the proteinase. Only a 3.2-fold purification was obtained in

these two steps together (Table lA), but the yield was high and they were required to improve the efficiency of the following purification steps. Since it was reported that HIV-l proteinase had a calculated isoelectric point of approximately 9.8 (40), a cation exchanger (CMSepharose) was chosen for the purification of the proteinase. At pH 6.5 all the proteinase was retained on the column. The enzyme was recovered with increasing concentrations of NaCl and found to elute at NaCl concentrations between 0.70 and 1.0 M (Fig. 3). This step resulted in a 30-fold purification and a 52% recovery (Table 1A). As a final purification step, size-exclusion chromatography was used (Fig. 4), taking advantage of the low molecular weight of the proteinase. Prior to size-exclu-

PURIFICATION S A

P

C

OF HIV-l

Ab

1

Ab

2

P

C

P

c

19

PROTEINASE TABLE

B

Purification

1

of HIV-1 Proteinase Total activity (nmol/minl

Total yield (%)

Total protein” bd

Specific activity* (nmol/min * mg)

Supernatant” Mesd CM-Sepharose’ Superose 12’

20 5.9 0.101

2.4 7.8 232

47 46 24

100 98 51

0.0087

1400

12

26

Supernatant’ Superose 12’

361 0.135

2.5 3000

903 406

45

Purification step A

B

FIG. 2. SDS-PAGE and immunoblot analysis of bacterial lysates. Bacterial lysates were subjected to SDS-PAGE and stained with Coomassie brilliant blue (A) or electrophoretically transferred to membranes and immunoblotted with anti-peptide antibodies (B). P represents lysates of bacteria expressing the proteinase gene; C is the corresponding control from bacteria transformed with the pRCB1 plasmid without the proteinase gene but expressing calbindin D,, (M, 8500); S is a molecular weight standard. Molecular weights of reference proteins are indicated in kilodaltons. Ab 1 and Ab 2 represent the two antibodies directed against peptides LDTGADDTVLEEMSLC and EEMSLPGRWKPKMIGC, respectively.

sion chromatography the active fractions from the CMSepharose column were pooled and concentrated by ultrafiltration, with a molecular weight cutoff filter of 5000, leaving the sample free of smaller proteins and peptides. At pH 5.5 the proteinase adsorbed to the Superose 12 column giving band broadening and poor recovery. At pH 7.0 the proteinase was eluted in earlier fractions which resulted in poor resolution and contamination with higher molecular weight proteins (data not shown). However, size-exclusion chromatography at pH 6.5 gave a g-fold purification and a 50% recovery (Table 1A). Due to the low protein concentration of the applied sample, only very small absorbances (~0.002 at 280 nm) could be monitored during elution. The proteinase was expected to elute at a position representing the dimer molecular weight, 20,000 Da, but it was retarded on the column to a position representing a protein of 5000 Da (18-19 ml). SDS-PAGE gels, however, showed that this protein had a molecular weight of 10,000 Da (Fig. 4). No proteinase was found in any other fraction, excluding the presence of higher molecular weight aggregates. A final yield of 26% of total expressed proteinase with a specific activity of 1400 nmol/min * mg was obtained with this purification method (Table 1A). The total amount of expressed proteinase was approximately 0.2% of total bacterial protein, and the proteinase was purified about 600-fold. Higher yield (45%) and specific activity (3000 nmol/min * mg) were obtained with the

100

Note. (A) Small-scale purification and (B) large-scale purification performed as described under Materials and Methods. Values have been corrected for portions of samples that could not be applied to columns due to the limited capacity in the small-scale purification; i.e., 1.5 ml of the 2 ml was applied to the CM-Sepharose column. a Protein concentration was determined with a Bio-Rad protein assay kit. * Activity was measured with the peptide substrate SQNYPIVQamide. ’ Supernatant from 100,OOOg centrifugation of 4.2 ml bacterial lysate for the small-scale purification and 50 ml for the large-scale purification, which correspond to approximately 1.2 and 14 g of bacterial pellet, respectively. d Ammonium sulfate precipitate dissolved in Mes buffer. e Pooled and concentrated fractions from cation-exchange column. ’ Pooled fractions from size-exclusion chromatography column.

large-scale purification (Table 1B) probably larger capacity of the columns used.

due to the

Characterization of the Purified Protein A determination of the five N-terminal amino acids of the purified HIV-l proteinase gave the expected se-

03 -

20 1 E ‘;

0,4 I

.

.E

g 0.3 -

B 5 10,

s * N ii 0.2-

I

0

I 0.6 Es g

!

5 l! 8

.*

0.2 p

2 4i

* 0.0

8 0.6 -

0.4

10,1-

a

1.0

0

k

20

40 Elution

volume

60

= 0.0

60

(ml)

FIG. 3. Cation-exchange chromatography of HIV-l proteinase. An ammonium sulfate precipitate was dissolved in Mes buffer and applied to a CM-Sepharose column (see text). The activity was measured using the peptide substrate SQNYPIVQ-amide as described under Materials and Methods.

20

GOOBAR s

14

33

34

35

36

37

38

39

40

ET AL. 12

3

4

5

6

7

a

9

10

36

5 -

FIG. 4. SDS-PAGE analysis of fractions from size-exclusion chromatography of HIV-1 proteinase. Pooled and concentrated material from cation-exchange chromatography was applied to a Superose 12 column. Eluted fractions were subjected to SDS-PAGE and silver stained as described under Materials and Methods. S is a molecular weight standard with molecular weights of 46,30,21,14.3,6.5, and 3.4 kDa; 33-40 represent fraction numbers. Proteolytic activity, measured with BSA as a substrate, was detected in fractions 34-38.

quence PQITL (Table 2), confirming that the nine extra amino acids upstream of the proteinase sequence had been cleaved from the original translation product and that the proteinase showed the expected autocatalytic activity (8,9). Proteolytic activity in samples of purified proteinase or in crude bacterial lysates expressing the HIV-l proteinase was measured by monitoring the cleavage of denatured bovine serum albumin by SDS-PAGE (Fig. 5), a method used to measure the proteolytic activity of avian myeloblastosis virus (AMV) proteinase (37). As for the AMV proteinase, cleavage of denatured BSA by the HIV-l proteinase was not observed at neutral pH, but at pH 5.5 cleavage of BSA was observed after 5 h incubation at 37°C (data not shown). For the analysis of different samples incubations were carried out overnight (16 h) at 37°C. The HIV-l proteinase activity was measured in the presence of 1 m M DTT and between 150 and 300 m M NaCl. The enzyme cleaved denatured BSA at multiple sites (Fig. 5). Three major bands were observed at molecular weights 60,000, 55,000, and

TABLE Amino-Terminal of Purified Cycle number

Amino acid

1 2 3 4 5

Pro Gln Ile Thr Leu

’ According

to Fig. 1

2

Sequence Analysis HIV-l Proteinase

Expected Pro Gln Ile Thr Leu

Yield (pmol) 23 19 13 11 15

FIG. 5. Proteolysis of bovine serum albumin by purified HIV-l proteinase. Proteinase was incubated with BSA, analyzed by SDSPAGE, and stained with Coomassie brilliant blue as described under Materials and Methods. Lane 1 is denatured BSA; lane 2 is a sample of the purified proteinase; lanes 3 and 4 are duplicates of incubations of denatured BSA with the active proteinase; lane 5 is an incubation of BSA with a heat-denatured proteinase sample; and lanes 6-10 are incubations of BSA and the active proteinase with increasing concentrations of pepstatin A, 10 nM, 100 nM, 1.0 PM, and 100 pM, respectively. Arrowheads indicate the main proteolytic products.

30,000. No cleavage of BSA was observed in control lysates lacking the proteinase gene incubated under the same conditions. Pepstatin A was found to inhibit cleavage of BSA by HIV-l proteinase completely, but at relatively high concentrations, alO0 PM (Fig. 5). This is in agreement with an earlier finding that pepstatin A is a poor inhibitor of the HIV-l proteinase, in comparison to other aspartic proteinases (40). To further characterize the catalytic properties of the purified proteinase, cleavage of the substrate peptide SQNYPIVQ-amide, representing the cleavage site between the matrix (~17) and the capsid (~24) proteins, was monitored by reversed-phase HPLC. This method was also used for the quantitative evaluation of the purification method (Table 1). The substrate peptide SQNYPIVQ-amide was eluted at a retention time of 12 min while the PIVQ-amide product eluted at 9 min (Fig. 6). An absorbance peak eluting at 8 min was identified as the other cleavage product SQNY, since it varied in size together with the PIVQ-amide product in a timeand concentration-dependent manner (data not shown). A small absorbance peak at 10 min was due to contamination in the substrate peptide preparation (Fig. 6). Cleavage of the peptide SQNYPIVQ-amide, at the subsaturating concentration of 0.4 mM, was linear with time for nearly 2 h at 37°C. Samples were incubated for 1 h at this temperature with a proteinase concentration between 60 and 200 nM. At this enzyme concentration, activity (1400 nmol/min . mg at 0.4 m M substrate, Table 1A) was directly proportional to substrate concentration up to 2 m M of substrate (data not shown), which gave a calculated V,,, of more than 7000 nmol/min * mg. A linear correspondence between amount of peptide

PURIFICATION

OF HIV-1 PROTEINASE

C

A

r1

FIG. 6. Proteolysis of peptide SQNYPIVQ-amide by HIV-l proteinase. Proteinase was incubated with the peptide and samples were analyzed by HPLC as described under Materials and Methods. Retention times are indicated in minutes. (A) Peptide incubated with heat-denatured proteinase; (B) peptide incubated with active proteinase; (C) mixture of peptide standards SQNYPIVQ-amide (12.01) and PIVQ-amide (9.25).

and integrated area was found up to at least 20 pg of peptide (data not shown). The amount of peptide measured was approximately 4 pugper sample. DISCUSSION

The aim of this work was to develop a highly efficient purification method for HIV-l proteinase expressed in E. coli from a synthetic gene. This method should fulfill two major demands, it should give high yields of proteinase with a high specific activity and include as few purification steps as possible. In order to obtain large amounts of proteinase, the gene was expressed in E. coli under conditions favorable for high expression of the synthetic gene. The amount of proteinase obtained was between 0.1 and 0.2 wt% of total E. coli protein (Table l), clearly much lower than that for calbindin Dgk (see C and P in Fig. 2A), which in this expression system reaches about 10 wt% (31). Also other differences were seen in the sets of proteins from bacteria expressing the two different plasmids (Fig. 2A). These differences increased with prolonged induction of proteinase expression, which caused bacteria to stop growing and die (data not shown). Thus the differences seen by SDS-PAGE between C and P in Fig. 2A are most probably due to proteinase activity. The lower expression of the proteinase compared to that of calbindin D,, is probably due to the shorter induction time; moreover proteinase expression from the plasmid may be limited very early by proteinase degradation of proteins essential for gene expression.

21

High yield and specific activity were obtained with a purification method that takes into consideration the optimal conditions for HIV-l proteinase activity and can be performed in one single buffer system to avoid dialysis and/or several precipitation steps. Several lines of evidence for the expression of the correct product by the bacteria were obtained. A protein of 10,000 Da was identified by SDS-PAGE and by immunoblot only in lysates of bacteria expressing the HIV-l proteinase gene. A mutant with the catalytically essential aspartic acid residue replaced by an asparagine was expressed but not processed to the native form. In addition, proteolytic activity measured with BSA or a peptide as substrate was detected only in bacteria expressing the proteinase gene. Two qualities of the HIV-l proteinase were used for its purification. These were its high isoelectric point and its small size. It is expected that lipids and carbohydrates, which could interfere with the purification, were removed in the initial centrifugation and precipitation steps. Virtually all proteinase was recovered from these precipitations without any loss in specific activity (see Results). The cation-exchange column used gave a very effective purification, but its small size/capacity allowed only 1.5 m l of the 2 m l from the previous step to be loaded. The small capacity of the CM-Sepharose column in this system may be due to the relatively low pH (6.5) used for a weak cation exchanger and the presence of residual amounts of ammonium sulfate in the applied sample, which was estimated not to exceed 50 mM. A 10 times larger column, however, allows the purification of at least 16 times larger sample volumes with improved efficiency (see Results). The large-scale purification shows that this procedures can be adapted for purification of large sample amounts with the same efficiency. Finally, the anomalous behavior of the proteinase in the size-exclusion chromatography column has been observed previously with another type of size-exclusion chromatography column (21) and may be due to interactions between the proteinase and the column matrix. This final purification step led to a preparation which gave a single band in silver-stained SDS electrophoresis gels (Fig. 4). During the course of this study several purification methods for the bacterially expressed HIV-l proteinase have been published (19-24); these give proteinase of purity comparable to ours (290%). These methods require at least one more chromatographic step and repeated dialysis or ammonium sulfate precipitations for buffer changes, or include the reversible denaturation of the proteinase, which usually results in lower specific activity due to the difficulties in obtaining a 100% yield of renaturation (23). Total yields of 215 pg proteinase per gram of bacterial protein (19) and 25 pg per 10 g of bacterial cell paste (20) have been reported. This is lower than the yield obtained with our purification

22

GOOBAR

method, which gives 26-45% of total expressed proteinase, or 328-374 pg proteinase per gram of bacterial protein. One purification method (21) was reported to give yields of 70% in a single purification step, but this was achieved after pooling five independent runs, and the amount of pure proteinase obtained per gram of bacterial protein was still lower (120 pg/gram bacterial protein) than that obtained with our purification method. An affinity purification method for the HIV-l proteinase also has been published (24); this involves two chromatographic steps and elution at high pH. The specific activity obtained was measured with a different substrate and it is therefore not possible to compare it with ours. The specific activity obtained with our purification method, measured with similar peptides coding for the same cleavage site, is much higher (1.4-3.0 pmol/ min * mg at 0.4 m M substrate) than that previously reported for the bacterially expressed proteinase (1.2 pmollmin. mg at 10 m M substrate) (19) and that reported for the chemically synthesized proteinase (0.7 pmol/min . mg at 1 m M substrate) (18). Finally, the cal7 pmol/min . mg is culated V,, value of approximately of the same order of magnitude as the previously reported value of 4.9 pmol/min . mg for a similar substrate (23).

ET AL. F. H. R., Chan, H. W., and Venkatesan, 3993-4002.

S. (1988) J. Virol. 62,

8. Debouck, C., Gorniak, J. G., Strickler, J. E., Meek, T. D., Metcalf, B. W., and Rosenberg, M. (1987) Proc. N&Z. Acad. Sci. USA 84, 8903-8906. 9. Giam, C-Z., and Boros, I. (1988) J. Biol. Chem. 263, 14,61714,620. 10. Mous, J., Heimer, E. P., and Le Grice, S. F. J. (1988) J. Virol. 62, 1433-1436. 11. Le Grice, S. F. J., Mills, J., and Mous, J. (1988) E M B O J. 7, 2547-2553. 12. Krausslich, H. G., Schneider, H., Zybarth, G., Carter, Wimmer, E. (1988) J. Viral. 62, 4393-4397.

C. A., and

13. Darke, P. L., Nutt, R. F., Brady, S. F., Garsky, V. M., Ciccarone, T. M., Leu, C-T., Lumma, P. K., Freidinger, R. M., Veber, D. F., and Sigal, I. S. (1988) Biochem. Biophys. Res. Commun. 156,297303. 14. Loeb, D. D., Hutchinson III, C. A., Edgell, M. H., Farmerie, W. G., and Swanstrom, R. (1989) J. Virol. 63, 111-121. 15. Kohl, N. E., Emini, E. A., Schleif, W. A., Davis, L. J., Heimbach, J. C., Dixon, R. A. F., Scolnick, E. M., and Sigal, I. S. (1988) Proc. Natl. Acad. Sci. USA 85,4686-4690. 16. Peng, C., Ho, B. K., Chang, T. W., and Chang, N. T. (1989) J. Viral. 63, 2550-2556. 17. Schneider, J., and Kent, S. B. H. (1988) Cell 54, 363-368. 18. Nutt, C. D., Sigal, Acad.

R. F., Brady, S. F., Darke, P. L., Ciccarone, T. M., Colton, Nutt, E. M., Rodney, J. A., Bennett, C. D., Waxman, L. H., I. S., Anderson, P. S., and Veber, D. F. (1988) Proc. Natl. Sci. USA 85, 7129-7133.

19. Darke, P. L., Leu, C-T.,Davis, L. J., Heimbach, J. C., Diehl, R.E., Hill, W. S., Dixon, R. A. F., and Sigal, I. S. (1989) J. Biol. Chem. 264,2307-2312.

ACKNOWLEDGMENTS This research was supported by grants from the Swedish National Board for Technical Development and by Symbicom AB. We thank Goran Utter for synthesizing the peptide used in the peptide cleavage assay, and Liselotte ohberg, Maryam Hessam Amiri, and Kerstin Enqvist for technical assistance. T. GrundstrBm was the recipient of a special researcher position from The Swedish Natural Science Research Council (Dnr 4892304).

REFERENCES 1. Ratner, L., Haseltine, W., Patarca, R., Livak, K. J., Starcich, B., Josephs, S. F., Doran, E. R., Rafalski, J. A., Whitehorn, E. A., Baumeister, K., Ivanoff, L., Petteway, S. R., Jr., Pearson, M. L., Lautenberger, J. A., Papas, T. S., Ghrayeb, J., Chang, N. T., Gallo, R. C., and Wong-Staal, F. (1985) Nature 313, 277-284. 2. Wain-Hobson, S., Sonigo, P., Danos, O., Cole, S., and Alizon, M. (1985) Cell 40,9-17. 3. Kramer, R. A., Schaber, M. D., Skalka, A. M., Ganguly, K., Wong-Staal, F., and Reddy, E. P. (1986) Science 231,1580-1584.

20. Krausslich, H. G., Ingraham, R. H., Skoog, M. T., Wimmer, E., Pallai, P. V., and Carter, C. A. (1989) Proc. Natl. Acad. Sci. USA 86,807-811. 21. Strickler, Magaard, 139-154.

J. E., Gorniak, V., Malinowski,

J., Dayton, B., Meek, T., Moore, M., J., and Debouk, C. (1989) Proteins 6,

22. Danley, D. E., Geoghegan, K. F., Scheld, K. G., Lee, S. E., Merson, J. R., Hawrylik, S. J., Rickett, G. A., Ammirati, M. J., and Hobart, P. M. (1989) Biochm. Biophys. Res. Commun. 165, 1043-1050. 23. Tomasselli, A. G., Olsen, M. K., Hui, J. O., Staples, D. J., Sawyer, T. K., Heinrikson, R. L., and Tomich, C-S. C. (1990) Biochemistry 29, 264-269. 24. Heimbach, J. C., Garsky, V. M., Michelson, S. R., Dixon, R. A. F., Sigal, I. S., and Darke, P. L. (1989) Biochem. Biophys. Res. Commun. 164,955-960. 25 Navia, M. A., Fitzgerald, P. M. D., McKeever, B. M., Leu, C-T., ’ Heimbach, J. C., Herber, W. K., Sigal, I. S., Darke, P. L., and Springer, J. P. (1989) Nature 337, 615-620.

6. Di Marzo Veronese, F., Copeland, T. D., Oroszlan, S., Gallo, R. C., and Sarngadharan, M. G. (1988) J. Virol. 62,795-801.

26. Wlodawer, A., Miller, M., Jaskblski, M., Sathyanarayana, B. K., Baldwin, E., Weber, I. T., Selk, L. M., Clawson, L., Schneider, J., and Kent, S. B. H. (1989) Science 245, 616-621. 27. Pearl, L. H., and Taylor, W. R. (1987) Nature 329, 351-354. 28. Meek, T. D., Dayton, B. D., Metcalf, B. W., Dreyer, G. B., Strickler, J. E., Gorniak, J. G., Rosenberg, M., Moore, M. L., Magaard, V. W., and Debouck, C. (1989) Proc. Natl. Acad. Sci. USA 86, 1841-1845.

7. Mervis, R. J., Ahmad, N., Lillehoj,

29. Lillehoj,

4. Graves, M. C., Lim, J. J., Heimer, E. P., and Kramer, Proc. Natl. Acad. Sci. USA 85, 2449-2453.

R. A. (1988)

5. Lightfoote, M. M., Coligan, J. E., Folks, T. M., Fauci, A. S., Martin, M. A., and Venkatesan, S. (1986) J. Viral. 60, 771-775.

E. P., Raum, M. G., Salazar,

E. P., Salazar, F. H. R., Mervis,

R. J., Raum, M. G.,

PURIFICATION Chan, H. W., Ahmad, N., and Venkatesan, 3053-3058.

OF HIV-l

S. (1988) J. Virol. 62,

30. Grundstrom, T., Zenke, W. M., Wintzenit, M., Matthes, H. W. D., Stauh, A., and Chambon, P. (1985) Nucleic Acid Res. 13, 33093316. 31. Brodin, P., Drakenberg, T., Thulin, E., For&n, S., and GrundStrom, T. (1989) Protein Eng. 2,353-358. 32. Bhatnagar, P. K., Papas, E., Blum, H. E., Milich, D. R., Nitecki, D., Karels, M. J., and Vyas, G. N. (1982) Proc. Nutl. Acad. Sci. USA 79,4400-4404. 33. Houghten, R. A. (1985) Proc. N&Z. Acud. Sci. USA 82, 51315135. 34. Schligger, H., and von Jagow, G. (1987) An&. Biochem. 166,369379.

23

PROTEINASE

35. Blum, H., Beier, H., and Gross, H. J. (1987) Ekctrophoresis 93-99. 36. Small, G. M., Imanaka, them. 169,405-409. 37. Dittmar,

T., and Lazarow,

K. J., and Moelling,

8,

P. B. (1988) Anal. Bio-

K. (1978) J. Viral. 28, 106-118.

38. Looman, A. C., Bodlaender, J., Cornstock, L. J., Eaton, D., Jhurani, P., de Boer, H. A., and von Knippenberg, P. H. (1987) EMBO J. 6, 2489-2492. T., Hofmann, T., Drakenberg, T., Thu39. Brodin, P., Grundstrom, lin, E., and For&n, S. (1986) Biochemistry 25,5371-5377. 40. Seelmeier, S., Schmidt, H., Turk, V., and von der Helm, K. (1988) Proc. Nutl. Acud. Sci. USA 85, 6612-6616.