SEARCH

SEARCH BY CITATION

Keywords:

  • Gram-positive bacterium;
  • Streptococcus pneumoniae;
  • tehB gene;
  • Tellurite resistance;
  • Filamentation

Abstract

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

Streptococcus pneumoniae is a Gram-positive bacterium which is naturally resistant to tellurite. In this study, we cloned and sequenced a homologue of the Escherichia coli tellurite resistance gene tehB from S. pneumoniae. It encoded a protein of 284 amino acids which is 86 residues longer than E. coli TehB, but similar in size to Haemophilus influenzae TehB and Eikenella corrodens hemagglutinin (Hag1) as well as homologues from Actinobacillus actinomycetemcomitans, Neisseria gonorrhoeae and Neisseria meningitidis. The S. pneumoniae TehB displayed 46–58% identity (52–68% similarity) to these proteins. The results in this study showed that the S. pneumoniae tehB alone not only conferred on E. coli high level resistance to tellurite, but also caused filamentous morphology in E. coli.


1Introduction

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

Tellurite resistance is uncommon in most Gram-negative bacteria. Early studies showed that tellurite resistance (TeR) was linked to the presence of plasmids of incompatibility (Inc) HI, HII, and P groups in bacteria [1, 2]. More recent studies have shown that various bacteria have chromosomal genes for TeR[3–6]. Some bacteria even have chromosomal homologues of plasmid TeR genes [3]. TeR determinants are very diverse and so far at least five different determinants have been characterized. Two determinants, designated terZABCDEF and kilAtelAB, were isolated from IncHI2 plasmid pMER610 (and R478) [7, 8] and IncPα plasmid RK2 [9], respectively. Another three determinants were identified in bacteria chromosomes. These include the E. coli tehAB, the Rhodobacter sphaeroides trgABcysK and telA, and the Pseudomonas syringae tpm gene [3]. There is no homology among these five determinants at either the DNA or protein level except for R. sphaeroides telA whose product is 65% similar to that of TelA of RK2 kilAtelAB. These genes, however, were not able to substitute for one another [3].

The precise mechanisms for TeR by these determinants are not known except for P. syringae tpm gene, whose product is 55% similar to human thiopurine methyltransferase and which could catalyze S-adenosyl-l-methionine (AdoMet) methylation of 6-mercaptopurine (a substrate for human thiopurine methyltransferase). It is possible that tellurite is methylated into volatile dimethyltelluride by this protein in vivo [6]. The E. coli TehB displays three conserved motifs (I, II and III) similar to those identified in many AdoMet-dependent non-nucleic acid methyltransferases, and site-directed mutagenesis has indicated that TehB may also be a methyltransferase [10]. Studies on the determinants terZABCDEF and kilAtelAB indicated that direct exclusion and efflux were not involved in resistance [3]. It was also noted that arsABC from the IncFI plasmid R773, which was originally found to be resistant to arsenic and antimony compounds, had an intermediate cross-resistance to tellurite medicated by an efflux pump mechanism [11].

In contrast to Gram-negative bacteria, molecular studies on TeR in Gram-positive bacteria have not been attempted, despite that fact that many Gram-positive bacteria, for example, Corynebacterium diphtheriae, Enterococcus faecalis and Staphylococcus aureus, are naturally resistant to tellurite [3]. In this study, we report that a single gene from human pathogen Streptococcus pneumoniae confers high level resistance to tellurite and causes filamentous morphology when expressed in E. coli.

2Materials and methods

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

2.1Bacterial strains and plasmids

Bacterial strains and the plasmids used in this study are listed in Table 1. E. coli JM109 was used as the cloning host. E. coli BL21(DE3)/pLysS was used for overexpression and for comparison of MICs. The S. pneumoniae type 1 strain was provided by M. Diadio, Department of Medical Microbiology and Immunology, University of Alberta. The plasmid pML204 is a construct containing the Eikenella corrodens hemagglutinin (hag1) gene recloned from the original plasmid pVKR204 provided by A. Progulske-Fox, University of Florida [14].

Table 1.  Bacterial strains, plasmids and oligonucleotide primers used in this study
Strain, plasmidDescriptionaSource/reference
  1. aApr, ampicillin resistance; Cmr, chloramphenicol resistance; Tcr, tetracycline resistance; hag, hemagglutinin gene; tehB, tellurite resistance gene B.

  2. bThe pLysS is a lysozyme-containing plasmid.

Strains  
JM109endA1 recA1 gyrA96 thi hsdR17 relA1 supE44Δ(lac-proAB)mcrA[FtraD36 proAB lacIqZΔM15][12]
Bl21(DE3)/pLysSbFompT hsdSB (rB mB) gal dcm (cIts857 indl Sam7nin5 lacUV5-T7gene1) pLysS (Cmr)Novagen, Madison, WI
Streptococcus pneumoniae type 1 M. Diadio
Plasmids  
pTZ18UPhagemid vector with lac and T7 promoters, AprBio-Rad, Hercules, CA
pTSPtehB-21.14-kb S. pneumoniae tehB fragment in pTZ18U, AprThis study
pTWT101E. coli tehAB fragment in pTZ19R, Apr[13]
pML111E. coli tehB in pTZ18U, AprThis study
pTWT134H. influenzae tehB homologue in pTZ18R, AprThis lab
pML204E. corrodens hag1 fragment in pTZ18U, Apr[14]; This study

2.2Media and growth conditions

The E. coli strains were cultured at 37°C in Luria-Bertani (LB) medium with or without 1.5% agar [15]. S. pneumoniae were cultured in Todd-Hewitt (Oxoid, Hampshire, UK) broth supplemented with 0.5% yeast extract (THY) or grown on THY-blood agar containing 5% defibrinated sheep blood (Dalynn, Calgary, Alta.), in a 5% CO2 atmosphere. Where appropriate, ampicillin was added to the medium at a concentration of 50 μg ml−1.

2.3Minimal inhibitory concentration (MIC)

The MICs for E. coli cultures containing various plasmids were determined using the agar dilution method as described previously [13]. The MIC for S. pneumoniae was determined by the broth dilution method [16]. Briefly, overnight cultures of S. pneumoniae were inoculated into THY tubes containing serial twofold dilutions of potassium tellurite, which were then kept in a 5% CO2 incubator overnight. The growth was measured by optical density at 600 nm. The lowest concentration of tellurite without growth was defined as the MIC.

2.4DNA techniques

Standard procedures for DNA manipulations were used as described previously [15]. Plasmid DNA was isolated from E. coli as described by Birnboim and Doly [17]. DNA fragments were purified from agarose gel with Spin-X centrifuge tube filters (Cornings, Cambridge, MA). Primers ML53 (5′-ATACTCGAGAATTCTTTTAGGACTTGCCAAA-3′) and ML54 (5′-ATACTGCAGGATCCTCTAACACATTTACCAA-3′) were designed based on the genome sequences of S. pneumoniae in NCBI database (underlined sequences indicate extensions introduced to create restriction sites XhoI and PstI, respectively). The S. pneumoniae tehB gene fragment (about 1.14 kb) was amplified by a modified colony PCR using primers ML53 and ML54 and cloned into pTZ18U (termed pTSPtehB-2). The sequence of this fragment was determined by using a thermal-sequencing radiolabeled terminator cycle sequencing kit purchased from Amersham (Cleveland, OH). A Perkin-Elmer DNA thermal cycler (model 480) (Norwalk, CT) was used for both tehB amplification and DNA sequencing. All enzymes in this study were obtained from Gibco BRL (Burlington, Ont.).

2.5Protein expression

Expression of the S. pneumoniae tehB gene was carried out as described previously [18]. BL21(DE3)/pLysS cells harboring various plasmids were grown in LB broth to an optical density of 0.5 at 600 nm. They were then washed and resuspended in supplemented M9 medium. After 1 h incubation at 37°C, the cells were induced by 0.1 mM IPTG for 30 min. Rifampicin was then added to a final concentration of 200 μg ml−1 for 30 min to stop E. coli RNA transcription. The [35S]methionine was subsequently added and incubation was continued for 1 h. Cells were then pelleted and resuspended cracking buffer. A 5-μl sample was loaded onto a 15% SDS-polyacrylamide gel. BenchMark Prestained Protein Ladder (Gibco BRL, Burlington, Ont.) was used as a standard marker.

2.6The accession number of the nucleotide sequence

The nucleotide sequence reported in this study has been deposited in GenBank with accession number AF079807. Sequence analysis and homology searches were performed with the software package of Genetics Computer Group (Madison, WI) and the NCBI server.

3Results

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

3.1Nucleotide sequence of the S. pneumoniae tehB gene

The complete S. pneumoniae tehB gene was amplified by PCR using primers MF53 and MF54, chosen because of homologies noted between E. coli tehAB[5] and S. pneumoniae DNA sequence. The resulting PCR fragments were cloned into pTZ18U. Four clones from three independent experiments were sequenced and found to have the same sequences as those that appeared in the S. pneumoniae in the NCBI unfinished genome database, except for one nucleotide change at 475 which resulted in an alteration of glutamine to glutamic acid. The open reading frame could encode a protein of 284 amino acids with a molecular mass of 32 444 Da.

3.2Comparison of the S. pneumoniae TehB and other homologous proteins

The deduced amino acid sequence of S. pneumoniae TehB was used to search for homologous proteins in databases. In addition to significant homology with TehB proteins of E. coli[4, 5] and H. influenzae[19], it also showed homologies with E. corrodens haemagglutinin (Hag1) [14] and homologous proteins from Actinobacillus actinomycetemcomitans, N. gonorrhoeae and N. meningitidis. Among these proteins, the E. coli TehB is the shortest, consisting of only 197 amino acids, all other proteins are 90–100 residues longer than that of E. coli at the N-terminus. The homology is distributed throughout the complete sequences, but the N-terminal extra part (with respect to E. coli TehB) showed a relatively lower degree of similarity (Fig. 2). Motifs I, II and III, which were found in many AdoMet-dependent non-nucleic acid methyltransferases, are also present in all these sequences (Fig. 1) [3, 10]. These homologies, determined as the percentage of identity and similarity between the amino acid sequences of the different proteins, are summarized in Table 2.

image

Figure 2. Expressions of S. pneumoniae TehB, E. coli TehB, H. influenzae TehB and E. corrodens Hag1. The proteins were synthesized and labeled as described in Section 2 and were separated on a 15% SDS-polyacrylamide gel. Lanes: 1, E. coli TehB; 2, E. corrodens Hag1; 3, H. influenzae TehB; 4, S. pneumoniae TehB; 5, plasmid pTZ18U only. TehBs of S. pneumoniae and H. influenzae and Hag1 of E. corrodens are indicated by a closed arrowhead, E. coli TehB by an open arrowhead. Molecular mass markers are indicated on the left.

Download figure to PowerPoint

image

Figure 1. Sequence comparison of S. pneumoniae TehB and homologues from six other bacteria. Sequences were aligned using the GCG software package (University of Wisconsin, Madison, WI). The identical residues across all sequences are highlighted in grey. Gaps (dots) are introduced to give the best alignment. SpTehB, S. pneumoniae TehB (accession number: AF079807, this study); HiTehB, H. influenzae TehB (U32827 [19]); EcHag1, E. corrodens hemagglutinin (Hag1) (P35647 [14]); EcTehB, E. coli TehB (M74072 [4]). Other sequences were from the NCBI unfinished genome sequences database.

Download figure to PowerPoint

Table 2.  Homologies of tellurite resistance TehB homologues from various bacteria
BacteriumIdentity (similarity)
 A. a.E. c.E. cor.H. i.N. g.N. m.
S. pneumoniae56 (65)57 (65)46 (52)58 (68)47 (54)47 (54)
A. actinomycetemcomitans 55 (62)44 (51)83 (95)44 (52)44 (53)
E. coli  51 (58)55 (61)49 (59)48 (59)
E. corrodens   45 (53)46 (56)46 (55)
H. influenzae    47 (56)47 (56)
N. gonorrhoeae     97 (98)

3.3The S. pneumoniae tehB confers E. coli tellurite resistance

S. pneumoniae encodes a tellurite MIC of 128 μg ml−1. The S. pneumoniae tehB was cloned into pTZ18U under both lacZ and T7 polymerase promoters (pTSPtehB-2). When this construct was transformed into JM109 which does not have a T7 polymerase gene, the JM109 strain encoded high level resistance to tellurite (128 μg ml−1), possibly due to the leaky expression from the lacZ promoter. Transformations of pTSPtehB-2 into BL21(DE3) which has a plac-controlled T7 polymerase gene were unsuccessful (even in the absence of IPTG), while other control plasmids were easily transformed. This indicated that S. pneumoniae TehB is toxic to E. coli cells; an excess of this protein may have resulted in lethality of BL21(DE3). This lethality could be circumvented by using a modified host BL21(DE3)/pLysS containing a lysozyme-producing plasmid (lysozyme is a natural inhibitor of T7 polymerase) as described below.

To study S. pneumoniae tehB toxicity and the tellurite resistance, we compared S. pneumoniae tehB and homologues from E. coli, H. influenzae as well as E. corrodens hag1 gene. All these genes were cloned into the same vector and transformed into BL21(DE3)/pLysS. When these transformants were not induced or induced with 0.1 mM of IPTG, pTSPtehB-2 conferred an MIC of 128–256 μg ml−1 similar to that which it conferred in JM109, other hosts into which tehB was transformed (pML111, pTWT134 and pML204) had an MIC of 64 μg ml−1. When induced with a higher concentration of IPTG (1 mM), however, all strains experienced a 2–4 fold decrease in tellurite resistance, except for the host carrying pTWT101 which contains the complete E. coli teh operon (including both tehA and tehB genes and the promoter region) (Table 3). The reason may be that pTWT101 uses its own promoter and the T7 promoter did not work on it effectively.

Table 3.  Tellurite resistancea in various TehB constructs in E. coli
ConstructGeneMIC (μg ml−1)
  00.11.0 (mM)b
  1. aMIC of potassium tellurite for S. pneumoniae is 128 μg ml−1.

  2. bIPTG concentration for induction in BL21(DE3)/pLysS.

  3. cComplete tehAB operon, including the promoter region.

pTZ18U− (vector)222
pTSPtehB-2S. pneumoniae tehB12825632–64
pTWT101cE. coli tehAB128128–256256
pML111E. coli tehB646432
pTWT134H. influenzae tehB646416–32
pML204E. corrodens hagl646416

3.4Filamentous morphology of E. coli caused by S. pneumoniae tehB gene

The E. coli JM109 harboring pTSPtehB-2 (under the lacZ promoter) was found to have a much slower growth rate and smaller colonies were produced. We examined cell morphology by light microscopy and found that cells were elongated. Scanning electron microscopy showed that these colonies consisting of long filaments. These filaments are typically over 5–10 times longer than the normal wild-type E. coli, shorter filaments (2–3 times longer) were rare (data not shown). JM109 containing H. influenzae tehB or E. corrodens hag1 also showed elongated morphology, but a little shorter than that caused by pTSPtehB (2–5 times longer). JM109 containing E. coli tehB has a normal morphology similar to that of JM109(pTWT101) (data not shown).

3.5Protein expression

The tehB genes were cloned under the T7 promoter and a BL21(DE3)/pLysS strain which carries plac-controlled T7 polymerase gene on the chromosome, was used for the expression (Fig. 2). The pTSPtehB-2 plasmid specified a polypeptide of 31 kDa, which is in agreement with the molecular mass of the deduced S. pneumoniae TehB (∼32.4 kDa). The H. influenzae TehB and E. corrodens Hag1, which have a similar length to that of S. pneumoniae TehB, showed similar sized protein bands (Fig. 2, closed arrowhead), whereas pTWT101 (and pML111) which have the short-form E. coli TehB displayed a band of ∼23 kDa, that is consistent with the previous expression result [4]. Control plasmid pTZ18U had no such bands (Fig. 2).

4Discussion

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

The S. pneumoniae tehB encode a protein of 284 amino acids, which has a high level of identity to the homologue from other bacteria (Table 2). The high-level identity among these bacteria is surprising when it is considered that E. corrodens Hag1 is a hemagglutinin [14]. Whether S. pneumoniae tehB also has erythrocyte-agglutinating activity is not known, however, E. corrodens hag1 has been demonstrated to confer tellurite resistance on E. coli in this study. This result highlights an interesting question: how is tellurite resistance associated with hemagglutination? Tellurite resistance might not be the main function for some of the TeR determinants, it may also encode additional properties which have not yet been identified [3]. The fact that tellurite resistance was frequently found to be linked with such phenotypes as bacteriophage inhibition and resistance to colicins supports this idea [2, 3, 8]. In this study, E. coli tehA, which is located in the same operon as tehB, was found to play a very limited role in tellurite resistance (Table 3), suggesting that tehAtehB may have other main function(s). The distribution of tehB homologues in different bacteria with a high-level homology may also indicate that TehB has other functions and is important for the survival of these bacteria.

Although in the E. coli K-12 genome, the tehA and tehB genes are located in one operon at 32.3 min [5], they are widely separated in H. influenzae[19]. To date 2.1 of the 2.2 Mb S. pneumoniae sequence has been completed and no tehA gene has been identified in the NCBI database. PCR analysis using primers based on E. coli tehA as well as Southern blotting with a tehA probe have not located a tehA gene (data not shown). These results, taken together with the tellurite MIC of 128 μg ml−1 specified by E. coli clones containing the cloned S. pneumoniae tehB gene, suggest that tehA is not required for TeR in S. pneumoniae.

E. coli filamentation caused by introduction of the S. pneumoniae tehB gene (as well as H. influenzae tehB and E. corrodens hag1) raises an interesting question, because two other plasmid-encoded TeR determinants, i.e. the ter operon from plasmid R478 [8] and the kil locus from RK2 [9], also caused similar filamentation [3]. No similarities were found among these TeR proteins or between these TeR proteins and known elements involved in cell division in E. coli. Other factors including E. coli AdoMet synthetase gene (metK) mutation [20] and E. coli elongation factor Tu (EF-Tu) alteration [21], also result in filamentation of the cells. How these mutations result in filamentation is not yet known. In the E. coli metK mutation, a low level of AdoMet is believed to affect alter cell morphology by either influencing methylation of EF-Tu, which associates with the cell membrane and could be involved in cell division, or alternatively it could influence some particular steps in cell division which need to be activated by methylation [20]. Like E. coli TehB, S. pneumoniae TehB have the three conserved motifs of AdoMet-dependent non-nucleic acid methyltransferases (Fig. 1). Therefore, E. coli TehB might function as a methyltransferase [10]. It is possible that overexpression of these proteins in E. coli exhausts the AdoMet pool to below a certain critical threshold, which may then affect the methylation of some cell division-related proteins resulting in filamentation. Future work is required to determine if this hypothesis is correct.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

We thank M. Diadio for kindly providing the S. pneumoniae strain and A. Progulske-Fox for the hag1-containing plasmid pVKR204. We are also grateful for R. Turner for cloning the H. influenzae tehB gene (pTWT134), M. Rooker for technical assistance, R. Sherburne for electron microscopy, Q. Jiang, D. Kelly and G. Wang for helpful discussion. This work was supported by the Medical Research Council of Canada (Grant MT 6200). D.E.T. is a Scientist with the Alberta Heritage Foundation for Medical Research.

References

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References
  • 1
    Summers, A.O. and Jacoby, G.A. (1977) Plasmid determined resistance and tellurite compounds. J. Bacteriol. 129, 276281.
  • 2
    Taylor, D.E. and Summers, A.O. (1979) Association of tellurium resistance and bacteriophage inhibition conferred by R plasmids. J. Bacteriol. 137, 14301433.
  • 3
    Taylor, D.E. (1999) Bacterial tellurite resistance. Trends Microbiol. 7, 111115.
  • 4
    Walter, E.G., Weiner, J.H. and Taylor, D.E. (1991) Nucleotide sequence and overexpression of the tellurite resistance determinant from the IncHII plasmid pHH1508a. Gene 101, 17.
  • 5
    Taylor, D.E., Hou, Y., Turner, R.J. and Weiner, J.H. (1994) Location of a potassium tellurite operon (tehAtehB) within the terminus of Escherichia coli K-12. J. Bacteriol. 176, 27402742.
  • 6
    Cournoyer, B., Watanabe, S. and Vivian, A. (1998) A tellurite-resistance genetic determinant from phytogenetic pseudomonads encodes a thiopurine methyltransferase. Biochim. Biophys. Acta 1379, 162168.
  • 7
    Jobling, M.G. and Ritchie, D.A. (1988) The nucleotide sequence of a plasmid determinant for resistance to tellurite anions. Gene 66, 245258.
  • 8
    Whelan, K.F., Colleran, E. and Taylor, D.E. (1995) Phage inhibition, colicin resistance, and tellurite resistance are encoded by a single cluster of genes on the IncHI2 plasmid R478. J. Bacteriol. 177, 50165027.
  • 9
    Walter, E.G., Thomas, C.M., Ibbotson, J.P. and Taylor, D.E. (1991) Transcriptional analysis, translational analysis, and the sequence of the kilA-tellurite resistance region of plasmid RK2 Ter. J. Bacteriol. 173, 11111119.
  • 10
    Liu, M. (1999) Studies of Tellurite Resistance Genes from Escherichia coli and Streptococcus pneumoniae. MSc Thesis, University of Alberta, Cagary, Alta.
  • 11
    Turner, R.J., Hou, Y., Weiner, J.H. and Taylor, D.E. (1992) The arsenical ATPase efflux pump mediates tellurite resistance. J. Bacteriol. 174, 30923094.
  • 12
    Yanisch-Perron, C., Vieira, J. and Messing, J. (1985) Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103119.
  • 13
    Turner, R.J., Taylor, D.E. and Weiner, J.H. (1997) Expression of Escherichia coli TehA gives resistance to antiseptics and disinfectants similar to that conferred by multidrug resistance efflux pumps. Antimicrob. Agents Chemother. 41, 440444.
  • 14
    Rao, V.K., Whitlock, J.A. and Progulske-Fox, A. (1993) Cloning, characterization and sequencing of two hemagglutinin genes from Eikenella corrodens. J. Gen. Microbiol. 139, 639650.
  • 15
    Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
  • 16
    Hindler, J. (1998) in: Essential Procedures for Clinical Microbiology (Isenberg, H.D., Ed.), pp. 205–254. ASM Press, Washington, DC.
  • 17
    Birnboim, H.C. and Doly, J. (1979) A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7, 15131523.
  • 18
    Ge, Z. and Taylor, D.E. (1996) Sequencing, expression and genetic characterization of the Helicobacter pylori ftsH gene encoding a protein homologous to members of a novel putative ATPase family. J. Bacteriol. 178, 61516157.
  • 19
    Fleischmann, R.D. et al. (1997) Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science 269, 495512.
  • 20
    Newman, E.B., Budman, L.I., Chan, E.C., Greene, R.C., Lin, R.T., Woldringh, C.L. and D'Ari, R. (1998) Lack of S-adenosylmethylthionine results in a cell division defect in Escherichia coli. J. Bacteriol. 180, 36143619.
  • 21
    Zeef, L.A., Mesters, J.R., Kraal, B. and Bosch, L. (1995) A growth-defective kirromycin-resistance EF-Tu Escherichia coli mutant and a spontaneously evolved suppression of the defect. Gene 165, 3943.