The central hydrophobic domain of the bovine papillomavirus E5 transforming protein can be functionally replaced by many hydrophobic amino acid sequences containing a glutamine.

The 44-amino-acid E5 transforming protein of bovine papillomavirus can induce growth transformation of cultured rodent fibroblast cell lines. Previous studies revealed that efficient transformation of mouse C127 cells by the E5 protein required a central core of hydrophobic amino acids and several specific carboxyl-terminal amino acids. Although a randomly derived sequence of hydrophobic amino acids could functionally replace the wild-type hydrophobic core, most such sequences could not. We show here that the conserved glutamine at position 17 in the hydrophobic domain is also important for transformation and that insertion of the glutamine can rescue the transforming activity of many but not all otherwise defective mutants containing random hydrophobic sequences. However, a class of mutants was identified that transform efficiently even in the absence of glutamine, demonstrating that the presence of this amino acid is not absolutely required for efficient transformation. E5 proteins containing the glutamine appear to display increased homodimer formation compared with mutant proteins lacking the glutamine, but this amino acid has no apparent effect on protein stability.     

Transforming activity of a 16-amino-acid segment of the bovine papillomavirus E5 protein linked to random sequences of hydrophobic amino acids.

The 44-amino-acid E5 protein of bovine papillomavirus type 1 is the smallest transforming protein yet described. Previous results from our laboratory indicate that a hydrophobic core and specific carboxyl-terminal amino acids are required for the E5 protein to exert its transforming function. In this study, additional substitution mutations were generated in the E5 gene to determine the minimal amino acid sequence requirements for focus formation in mouse C127 cells. In most cases examined, substitution of the hydrophobic middle third of the E5 protein with unrelated hydrophobic sequences severely inhibited transforming activity. However, we have identified one hydrophobic amino acid sequence apparently unrelated to the wild-type one that can replace the middle third of the wild-type E5 protein without affecting the ability of the protein to stably transform cells or interact with cell membranes. Furthermore, a mutant E5 protein in which only the carboxyl-terminal 16 amino acids of the protein have been derived from E5 sequences retains transforming activity. Since several residues in the carboxyl-terminal portion of the E5 protein can be freely substituted with different amino acids (B. H. Horwitz, A. L. Burkhardt, R. Schlegel, and D. DiMaio, Mol. Cell. Biol. 8:4071-4078, 1988), the results reported here imply that much of the specific information necessary for cell transformation can be supplied by a subset of the carboxyl-terminal 16 amino acids of this protein.     

Specific locations of hydrophilic amino acids in constructed transmembrane ligands of the platelet-derived growth factor beta receptor

The 44 amino acid E5 transmembrane protein is the primary oncogene product of bovine papillomavirus. Homodimers of the E5 protein activate the cellular PDGF beta receptor tyrosine kinase by binding to its transmembrane domain and inducing receptor dimerization, resulting in cellular transformation. To investigate the role of transmembrane hydrophilic amino acids in receptor activation, we constructed a library of dimeric small transmembrane proteins in which 16 transmembrane amino acids of the E5 protein were replaced with random, predominantly hydrophobic amino acids. A low level of hydrophilic amino acids was encoded at each of the randomized positions, including position 17, which is an essential glutamine in the wild-type E5 protein. Library proteins that induced transformation in mouse C127 cells stably bound and activated the PDGF beta receptor. Strikingly, 35% of the transforming clones had a hydrophilic amino acid at position 17, highlighting the importance of this position in activation of the PDGF beta receptor. Hydrophilic amino acids in other transforming proteins were found adjacent to position 17 or at position 14 or 21, which are in the E5 homodimer interface. Approximately 22% of the transforming proteins lacked hydrophilic amino acids. The hydrophilic amino acids in the transforming clones appear to be important for driving homodimerization, binding to the PDGF beta receptor, or both. Interestingly, several of the library proteins bound and activated PDGF beta receptor transmembrane mutants that were not activated by the wild-type E5 protein. These experiments identified transmembrane proteins that activate the PDGF beta receptor and revealed the importance of hydrophilic amino acids at specific positions in the transmembrane sequence. Our identification of transformation-competent transmembrane proteins with altered specificity suggests that this approach may allow the creation and identification of transmembrane proteins that modulate the activity of a variety of receptor tyrosine kinases.     

A glutamine residue in the membrane-associating domain of the bovine papillomavirus type 1 E5 oncoprotein mediates its binding to a transmembrane component of the vacuolar H(+)-ATPase.

The 44-amino-acid E5 oncoprotein is the major transforming protein of bovine papillomavirus type 1. It is a highly hydrophobic polypeptide which dimerizes and localizes to the Golgi apparatus and endoplasmic reticulum membranes. Recent evidence suggests that E5 modulates the phosphorylation and internalization of the epidermal growth factor and colony-stimulating factor 1 receptors and constitutively activates platelet-derived growth factor receptors in C127 and FR3T3 cells. Although no direct interaction with these growth factor receptors has yet been identified, the E5 oncoprotein has been shown recently to interact with the hydrophobic 16-kDa component of the vacuolar H(+)-ATPase (16K protein) [D. J. Goldstein, M. E. Finbow, T. Andresson, P. McLean, K. Smith, V. Bubb, and R. Schlegel, Nature (London) 352:347-349, 1991]. In the current study, we have further analyzed the E5-16K protein complex by fast protein liquid chromatography and shown that each E5 dimer appears to bind two 16K proteins. In order to define the specific amino acid residues of E5 which participate in this binding, mutated E5 epitope fusion proteins were analyzed for their ability to coprecipitate 16K protein. Transformation-defective mutants containing amino acid substitutions within the short hydrophilic carboxyl-terminal domain retained the ability to associate with the 16K protein. However, E5 mutants lacking the glutamine residue in the hydrophobic domain were markedly inhibited in 16K protein binding. Most interestingly, the placement of a glutamine in several random hydrophobic sequences facilitated 16K protein binding, defining this residue as a potential binding site for the 16K protein component of the proton pump and exemplifying the critical role of hydrophilic amino acids for mediating specific interactions between transmembrane proteins.     

A point mutational analysis of human papillomavirus type 16 E7 protein.

The E7 open reading frame of human papillomavirus type 16 (HPV16) has been shown to be selectively retained in cervical tumors and to encode both transforming and trans-activating functions in murine cells, supporting the notion that expression of E7 contributes towards the progression of premalignant cervical lesions. A comparison among E7 sequences of different HPV types reveals some homology at the amino acid level. Of particular interest are two regions, one which contains significant homology to a region of adenovirus E1a and simian virus 40 large T (LT), and a second region which contains two conserved Cys-X-X-Cys motifs. To determine the importance of these domains to the function of the E7 protein, a series of mutants carrying substitutions at amino acids in the region of E1a-LT homology and at the Cys-X-X-Cys motifs were constructed. The mutated E7 sequences were placed under the control of a strong heterologous promoter (Moloney long terminal repeat), and the activity of the mutants was assayed in NIH 3T3 cells, a cell line in which both the transforming function and the trans-activating function of E7 could be determined. A single amino acid substitution analogous to a mutation in E1a which destroys the transforming ability of this protein abolished both transformation and trans-activation by E7. Mutations at the Cys-X-X-Cys motifs demonstrated that this region contributes to the transforming potential of E7, although proteins in which both motifs were interrupted retained a low level of transforming activity. Mutations in the region of E1a-LT homology which occur within a recognition sequence for casein kinase II did not markedly affect transforming activity of E7 but severely reduced trans-activating ability. This indicates that efficient trans-activation is not required for transformation by HPV16 E7 in these cells.     

Bovine papillomavirus type 1 oncoprotein E5 stimulates the utilization of superoxide radicals in the mouse fibroblast cell line C127

The major transforming protein of bovine papillomavirus type 1 (BPV-1) is a small hydrophobic polypeptide, the E5 gene product, localized in the cellular membranes and modulating various pathways in the cell. Many studies have shown that reactive oxygen species (ROS) are essential in several biological processes, including cell transformation by oncogenes, but unregulated ROS are highly toxic to cells. We studied the effect of the bovine papillomavirus protein E5 and its mutants on the level of the superoxide radicals in the mouse fibroblast cell line C127. The superoxide level in C127 cells transfected with the E5-expressing plasmids were measured by nitroblue tetrazolium reduction. Relative concentrations of intracellular peroxide were determined by using 2,7-dichlorofluorescin diacetate. Our results showed that all transforming mutants of E5 reduced the level of superoxide in C127 cells, besides the activity of superoxide dismutase (SOD) and level of peroxides was not altered. In the presence of neopterin, an inhibitor of the superoxide-producing enzymes, the reduction of superoxide level correlated with the transforming ability of the E5-mutants. The inhibitor of the protein tyrosine kinase, tyrphostin 25 and inhibitors of oxygenases of the arachidonic acid metabolism, aspirin and nordihydroguaiaretic acid, blocked the effect of BPV-1 E5. We conclude that BPV-1 E5 and its transforming mutants are able to modulate the level of superoxide and stimulate the utilization of superoxide through protein tyrosine kinases and oxygenases of the arachidonic acid metabolism.     

Identification of separate domains in the adenovirus E1A gene for immortalization activity and the activation of virus early genes.

The transformation and early adenovirus gene transactivation functions of the E1A region were analyzed with deletion and point mutations. Deletion of amino acids from position 86 through 120 had little effect on the lytic or transforming functions of the E1A products, while deletion of amino acids from position 121 through 150 significantly impaired both functions. The sensitivity of the transformation function to alterations in the region from amino acid position 121 to 150 was further indicated by the impairment of transforming activity resulting from single amino acid substitutions at positions 124 and 135. Interestingly, conversion of a cysteine residue at position 124 to glycine severely impaired the transformation function without affecting the early adenovirus gene activating functions. Single amino acid substitutions in a different region of the E1A gene had the converse effect. All the mutants produced polypeptides of sufficient stability to be detected by Western immunoblot analysis. The single amino acid substitutions at positions 124 and 135, although impairing the transformation functions, did not detectably alter the formation of the higher-apparent-molecular-weight forms of the E1A products.     

Mutational analysis of bovine papillomavirus type 1 E5 peptide domains involved in induction of cellular DNA synthesis.

Early gene E5 of bovine papillomavirus type 1 encodes a 44-amino-acid protein whose expression can transform immortalized mouse cell lines. We have previously reported that a chemically synthesized E5 peptide functions to induce cellular DNA synthesis upon microinjection into growth-arrested mouse cells. We further defined the two E5 domains essential for the full DNA synthesis induction activity by the analysis of E5 deletion and amino acid substitution mutant peptides. The first domain is the C-terminal 13-amino-acid core which is sufficient to activate DNA synthesis at high peptide concentration and contains two essential, highly conserved cysteine residues. The second domain is the 7-amino-acid hydrophobic sequence contiguous to the core domain which is sufficient to confer a 1,000-fold higher molar specific activity to the E5 peptide. A random hydrophobic sequence, but not charged amino acids, fulfills the function of the second domain.     

Mutational Analysis of the Fusion Peptide of Moloney Murine Leukemia Virus Transmembrane Protein p15E

Fusion peptides are hydrophobic sequences located at the N terminus of the transmembrane (TM) envelope proteins of the orthomyxoviruses and paramyxoviruses and several retroviruses. The Moloney murine leukemia virus TM envelope protein, p15E, contains a hydrophobic stretch of amino acids at its N terminus followed by a region rich in glycine and threonine residues. A series of single amino acid substitutions were introduced into this region, and the resulting proteins were examined for their abilities to be properly processed and transported to the cell surface and to induce syncytia in cells expressing the ecotropic receptor. One substitution in the hydrophobic core and several substitutions in the glycine/threonine-rich region that prevented both cell-cell fusion and the transduction of NIH 3T3 cells when incorporated into retroviral vector particles were identified. In addition, one mutation that enhanced the fusogenicity of the resulting envelope protein was identified. The fusion-defective mutants trans dominantly interfered with the ability of the wild-type envelope protein to cause syncytium formation in a cell-cell fusion assay, although no trans-dominant inhibition of transduction was observed. Certain substitutions in the hydrophobic core that prevented envelope protein processing were also found. These data indicate that the N-terminal region of p15E is important both for viral fusion and for the correct processing and cell surface expression of the viral envelope protein.     

Two regions of the adenovirus early region 1A proteins are required for transformation.

Regions of the adenovirus type 5 early region 1A (E1A) proteins that are required for transformation were defined by using a series of deletion mutants. Deletion mutations collectively spanning the entire protein-coding region of E1A were constructed and assayed for their ability to cooperate with an activated ras oncogene to induce transformation in primary baby rat kidney cells. Two regions of E1A (amino acids 1 to 85 and 121 to 127) were found to be essential for transformation. Deletion of all or part of the region from amino acids 121 to 127 resulted in a total loss of transforming ability. An adjacent stretch of amino acids (residues 128 to 139), largely consisting of acidic residues, was found to be dispensable for transformation but appeared to influence the efficiency of transformation. Amino acids 1 to 85 made up a second region of the E1A protein that was essential for transformation. Deletion of all or part of this region resulted in a loss of the transforming activity. Even a mutation resulting in a single amino acid change at position 2 of the polypeptide chain was sufficient to eliminate transformation. Deletion of amino acids 86 to 120 or 128 to 289 did not eliminate transformation, although some mutations in these regions had lowered efficiencies of transformation. Foci induced by transformation-competent mutants could be expanded into cell lines that retained their transformed morphology and constitutively expressed the mutant E1A proteins.     

Conserved arginine residue in the membrane-spanning domain of HIV-1 gp41 is required for efficient membrane fusion

Despite the high mutation rate of HIV-1, the amino acid sequences of the membrane-spanning domain (MSD) of HIV-1 gp41 are well conserved. Arginine residues are rarely found in single membrane-spanning domains, yet an arginine residue, R⁶⁹⁶ (the numbering is based on that of HXB2), is highly conserved in HIV-1 gp41. To examine the role of R⁶⁹⁶, it was mutated to K, A, I, L, D, E, N, and Q. Most of these substitutions did not affect the expression, processing or surface distribution of the envelope protein (Env). However, a syncytia formation assay showed that the substitution of R⁶⁹⁶ with amino acid residues other than K, a naturally observed mutation in the gp41 MSD, decreased fusion activity. Substitution with hydrophobic amino acid residues (A, I, and L) resulted in a modest decrease, while substitution with D or E, potentially negatively-charged residues, almost abolished the syncytia formation. All the fusion-defective mutants showed slower kinetics with the cell-based dual split protein (DSP) assay that scores the degree of membrane fusion based on pore formation between fusing cells. Interestingly, the D and E substitutions did show some fusion activity in the DSP assays, suggesting that proteins containing D or E substitutions retained some fusion pore-forming capability. However, nascent pores failed to develop, due probably to impaired activity in the pore enlargement process. Our data show the importance of this conserved arginine residue for efficient membrane fusion.     

Expression and mutational analysis of Autographa californica nucleopolyhedrovirus HCF-1: functional requirements for cysteine residues

The host cell-specific factor 1 gene (hcf-1) of the baculovirus Autographa californica multiple nucleopolyhedrovirus is required for efficient virus growth in TN368 cells but is dispensable for virus replication in SF21 cells. However, the mechanism of action of hcf-1 is unknown. To begin to understand its function in virus replication we have investigated the expression and localization pattern of HCF-1 in infected cells. Analysis of virus-infected TN368 cells showed that hcf-1 is expressed at an early time in the virus life cycle, between 2 and 12 h postinfection, and localized the protein to punctate nuclear foci. Through coprecipitation experiments we have confirmed that HCF-1 self-associates into dimers or higher-order structures. We also found that overexpression of HCF-1 repressed expression from the hcf-1 promoter in transient reporter assays. Mutagenesis of cysteine residues within a putative RING finger domain in the amino acid sequence of HCF-1 abolished self-association activity and suggests that the RING domain may be involved in this protein-protein interaction. A different but overlapping set of cysteine residues were required for efficient gene repression activity. Functional analysis of HCF-1 mutants showed that the cysteine amino acids required for both self-association and gene repression activities of HCF-1 were also required for efficient late-gene expression and occlusion body formation in TN368 cells. Mutational analysis also identified essential charged and hydrophobic amino acids located between two of the essential cysteine residues. We propose that HCF-1 is a RING finger-containing protein whose activity requires HCF-1 self-association and gene repression activity.     

The transmembrane domain of the E5 oncoprotein contains functionally discrete helical faces

The E5 protein of bovine papillomavirus is a 44-amino acid, Golgi-resident, type II transmembrane protein that efficiently transforms immortalized mouse fibroblasts. The transmembrane (TM) domain of E5 is not only critical for biological activity, it also regulates interactions with cellular targets including the platelet derived growth factor receptor (PDGF-R) and the 16-kDa subunit of the vacuolar proton ATPase (V-ATPase). In order to define the specific TM amino acids essential for E5 biological and biochemical activity, we performed scanning alanine mutagenesis on 25 of the 30 potential TM residues and genetically mapped discrete alpha-helical domains which separately regulated the ability of E5 to bind PDGF-R, activate PDGF-R, and to form oligomers. Alanine substitutions at positions 17, 21, and 24 (which lie on the same helical face) greatly inhibited E5 association with the PDGF-R, suggesting that this region comprises the receptor binding site. PDGF-R activation also mapped to a specific but broader domain in E5; mutant proteins with alanines on one helical face (positions 8, 9, 11, 16, 19, 22, and 23) continued to induce PDGF-R tyrosine phosphorylation, whereas mutant proteins with alanines on the opposite helical face (positions 7, 10, 13, 17, 18, 21, 24, and 25) did not, indicating that the latter helical face was critical for mediating receptor transphosphorylation. Interestingly, these "activation-defective" mutants segregated into two classes: 1) those that were unable to form dimers but that could still form higher order oligomers and transform cells, and 2) those that were defective for PDGF-R binding and were transformation-incompetent. These findings suggest that the ability of E5 to dimerize and to bind PDGF-R is important for receptor activation. However, since several transformation-competent E5 mutants were defective for binding and/or activating PDGF-R, it is apparent that E5 must have additional activities to mediate cell transformation. Finally, alanine substitutions also defined two separate helical faces critical for E5/E5 interactions (homodimer formation). Thus, our data identify distinct E5 helical faces that regulate homologous and heterologous intramembrane interactions and define two new classes of biologically active TM mutants.     

Two-Dimensional Polyacrylamide Gel Analysis of Plodia interpunctella Granulosis Virus

Polyomavirus middle-T antigen contains a contiguous sequence of 22 hydrophobic amino acids near the carboxyl terminus, which is the putative membrane-binding domain of the protein. The DNA encoding this region was mutated to form a series of deletions, insertions, and substitutions called RX mutants. The phenotypes of these mutants fall into three groups based on the transforming and biochemical properties of their encoded proteins. The first group, with deletions outside but proximal to the hydrophobic domain, displayed an essentially wild-type phenotype. A second group, with extensive deletions within the region encoding the hydrophobic domain, expressed middle-T species which did not fractionate with cellular membranes or associate with pp60c-src and which were defective in their ability to transform. A third group of mutants with more subtle predicted alterations in the hydrophobic domain were wild type for the biochemical parameters investigated but were unable to transform cultured rodent cells. These observations are consistent with previous findings that membrane association plays an important role in transformation by middle-T and that, whereas association between middle-T and pp60c-src is a necessary correlate of transformation, it is not sufficient. A comparison of murine polyomavirus middle-T and a newly described hamster papovavirus putative middle-T revealed a strong homology between their respective hydrophobic-domain amino acid sequences. This homology is not observed in the anchorage domains of other model proteins, and this may imply that the middle-T hydrophobic domain is important in transformation for reasons other than simple membrane association.     

The BPV-1 E5 protein, the 16 kDa membrane pore-forming protein and the PDGF receptor exist in a complex that is dependent on hydrophobic transmembrane interactions.

The E5 oncoprotein of bovine papillomavirus type 1 is a 44 amino acid, highly hydrophobic protein that induces the stable transformation of immortalized murine fibroblasts, presumably through its activation of growth factor receptors. Previous studies have shown that the E5 protein complexes with the 16 kDa (16k) pore-forming protein of vacuolar H(+)-ATPases. This integral membrane protein is essential for the acidification and function of subcellular compartments that process growth factor receptors. Using an SV40 expression system in COS cells, we analyzed whether the E5-16k complexes bind additional cellular proteins, including growth factor receptors. These studies demonstrate that E5 binds to both the 16k protein and the PDGF receptor and that this tri-component complex can be isolated with antibodies specific for each protein. Importantly, the 16k protein bound to the PDGF receptor in the absence of E5, suggesting that E5 binds to the PDGF receptor via its interaction with the 16k protein. An E5 mutant lacking the hydrophilic carboxyl-terminal 14 amino acids retained binding to both 16k and the PDGF receptor, indicating that E5 binds to these proteins through its hydrophobic, membrane-associating domain. These studies reveal that hydrophobic, intramembrane interactions govern the association of E5, 16k and the PDGF receptor, suggesting a ligand-independent mechanism for receptor activation and a potential link between receptor signal transduction pathways and membrane pore activity.     

Identification of amino acids in the transmembrane and juxtamembrane domains of the platelet-derived growth factor receptor required for productive interaction with the bovine papillomavirus E5 protein.

The bovine papillomavirus E5 protein forms a stable complex with the cellular platelet-derived growth factor (PDGF) beta receptor, resulting in receptor activation and cell transformation. Amino acids in both the putative transmembrane domain and extracytoplasmic carboxyl-terminal domain of the E5 protein appear important for PDGF receptor binding and activation. Previous analysis indicated that the transmembrane domain of the receptor was also required for complex formation and receptor activation. Here we analyzed receptor chimeras and point mutants to identify specific amino acids in the PDGF beta receptor required for productive interaction with the E5 protein. These receptor mutants were analyzed in murine Ba/F3 cells, which do not express endogenous receptor. Our results confirmed the importance of the transmembrane domain of the receptor for complex formation, receptor tyrosine phosphorylation, and mitogenic signaling in response to the E5 protein and established that the threonine residue in this domain is required for these activities. In addition, a positive charge in the extracellular juxtamembrane domain of the receptor was required for E5 interaction and signaling, whereas replacement of the wild-type lysine with either a neutral or acidic amino acid inhibited E5-induced receptor activation and transformation. All of the receptor mutants defective for activation by the E5 protein responded to acute treatment with PDGF and to stable expression of v-Sis, a form of PDGF. The required juxtamembrane lysine and transmembrane threonine are predicted to align precisely on the same face of an alpha helix packed in a left-handed coiled-coil geometry. These results establish that the E5 protein and v-Sis recognize distinct binding sites on the PDGF beta receptor and further clarify the nature of the interaction between the viral transforming protein and its cellular target.     

An oligomeric form of simian virus 40 large T-antigen is immunologically related to the cellular tumor antigen p53.

The induction of tumors and cellular transformation mediated by polyomavirus requires the action of middle T antigen. Accordingly, we have begun to define the domains of the viral protein important for these processes to learn more about its site and mechanism of action. One of the domains of middle T antigen which is thought to be important for its function includes a stretch of acidic amino acids and a vicinal tyrosine residue (tyrosine 315), the major site of tyrosine phosphorylation in vitro. To determine whether these acidic amino acids and tyrosine 315 are required to maintain the transforming activity of middle T antigen, we constructed deletions within the DNA sequences encoding these amino acids and measured the capacity of the resulting mutants to transform Rat-1 cells in culture. This was accomplished by using in vitro mutagenesis techniques with molecularly cloned polyomavirus DNA. Seven mutants were isolated. Five of these proved incapable of transforming Rat-1 cells and were found to contain deletions which altered the reading frame for middle T antigen. However, two mutants, pPdl1-4 and pPdl2-7, retained the capacity to transform Rat-1 cells at high frequencies. The middle T antigen encoded by one of these mutants, pPdl1-4, lacks part of the acidic string of amino acids but not tyrosine 315 (amino acids 304 through 310 are deleted), whereas the middle T antigen encoded by the other mutant, pPdl2-7, lacks the entire acidic amino acid stretch as well as tyrosine 315 (amino acids 285 through 323 are deleted). Rat-1 cells transformed by one or the other mutant DNA displayed a fully transformed phenotype, including the capacity to form tumors in animals. These results prove that the major site of tyrosine phosphorylation in middle T antigen and the acidic amino acids which precede it are not essential for its transforming activity.     

Fine structure analysis of a protein folding transition state; distinguishing between hydrophobic stabilization and specific packing

Developing a detailed understanding of the structure and energetics of protein folding transition states is a key step in describing the folding process. The phi-value analysis approach allows the energetic contribution of side-chains to be mapped out by comparing wild-type with individual mutants where conservative changes are introduced. Studies where multiple substitutions are made at individual sites are much rarer but are potentially very useful for understanding the contribution of each element of a side-chain to transition state formation, and for distinguishing the relative importance of specific packing versus hydrophobic interactions. We have made a series of conservative mutations at multiple buried sites in the N-terminal domain of L9 in order to assess the relative importance of specific side-chain packing versus less specific hydrophobic stabilization of the transition state. A total of 28 variants were prepared using both naturally occurring and non-naturally occurring amino acids at six sites. Analysis of the mutants by NMR and CD showed no perturbation of the structure. There is no correlation between changes in hydrophobicity and changes in stability. In contrast, there is excellent linear correlation between the hydrophobicity of a side-chain and the log of the folding rate, ln(k(f)). The correlation between ln(k(f)) and the change in hydrophobicity holds even for substitutions that change the shape and/or size of a side-chain significantly. For most sites, the correlation with the logarithm of the unfolding rate, ln(k(u)), is much worse. Mutants with more hydrophobic amino acid substitutions fold faster, and those with less hydrophobic amino acid substitutions fold slower. The results show that hydrophobic interactions amongst core residues are an important driving force for forming the transition state, and are more important than specific tight packing interactions. Finally, a number of substitutions lead to negative phi-values and the origin of these effects are described.