Dissecting the clinicopathologic, genomic, and immunophenotypic correlates of KRASG12D-mutated non-small-cell lung cancer

BACKGROUND

Allele-specific KRAS inhibitors are an emerging class of cancer therapies. KRAS-mutant (KRASMUT) non-small-cell lung cancers (NSCLCs) exhibit heterogeneous outcomes, driven by differences in underlying biology shaped by co-mutations. In contrast to KRASG12C NSCLC, KRASG12D NSCLC is associated with low/never-smoking status and is largely uncharacterized.

PATIENTS AND METHODS

Clinicopathologic and genomic information were collected from patients with NSCLCs harboring a KRAS mutation at the Dana-Farber Cancer Institute (DFCI), Memorial Sloan Kettering Cancer Center, MD Anderson Cancer Center, and Imperial College of London. Multiplexed immunofluorescence for CK7, programmed cell death protein 1 (PD-1), programmed death-ligand 1 (PD-L1), Foxp3, and CD8 was carried out on a subset of samples with available tissue at the DFCI. Clinical outcomes to PD-(L)1 inhibition ± chemotherapy were analyzed according to KRAS mutation subtype.

RESULTS

Of 2327 patients with KRAS-mutated (KRASMUT) NSCLC, 15% (n = 354) harbored KRASG12D. Compared to KRASnon-G12D NSCLC, KRASG12D NSCLC had a lower pack-year (py) smoking history (median 22.5 py versus 30.0 py, P < 0.0001) and was enriched in never smokers (22% versus 5%, P < 0.0001). KRASG12D had lower PD-L1 tumor proportion score (TPS) (median 1% versus 5%, P < 0.01) and lower tumor mutation burden (TMB) compared to KRASnon-G12D (median 8.4 versus 9.9 mt/Mb, P < 0.0001). Of the samples which underwent multiplexed immunofluorescence, KRASG12D had lower intratumoral and total CD8+PD1+ T cells (P < 0.05). Among 850 patients with advanced KRASMUT NSCLC who received PD-(L)1-based therapies, KRASG12D was associated with a worse objective response rate (ORR) (15.8% versus 28.4%, P = 0.03), progression-free survival (PFS) [hazard ratio (HR) 1.51, 95% confidence interval (CI) 1.45-2.00, P = 0.003], and overall survival (OS; HR 1.45, 1.05-1.99, P = 0.02) to PD-(L)1 inhibition alone but not to chemo-immunotherapy combinations [ORR 30.6% versus 35.7%, P = 0.51; PFS HR 1.28 (95%CI 0.92-1.77), P = 0.13; OS HR 1.36 (95%CI 0.95-1.96), P = 0.09] compared to KRASnon-G12D.

CONCLUSIONS

KRASG12D lung cancers harbor distinct clinical, genomic, and immunologic features compared to other KRAS-mutated lung cancers and worse outcomes to PD-(L)1 blockade. Drug development for KRASG12D lung cancers will have to take these differences into account.

Abstract P1-05-11: Comprehensive comparison of breast cancer molecular portraits by African and European ancestry in the cancer genome atlas

Background: African American breast cancer patients have worse survival rates than European American patients. Although racial differences in the distribution of breast cancer intrinsic subtype are known, it is unclear if there are other inherent genomic differences contributing to this racial outcome disparity.

Methods: We defined patient race based on genomic ancestry and compared multiple molecular features of breast cancer between 154 black and 776 white patients in The Cancer Genome Atlas (TCGA). We examined the contribution of these molecular features to survival outcomes using Cox proportional hazards models. We also estimated the heritability of breast cancer subtypes using a mixed effect model.

Results: Compared to whites, black patients had higher odds of basal-like (odds ratio=3.80, p&lt;0.001) and HER2-enriched (odds ratio=2.22, p=0.027) breast cancers in reference to luminal A subtype. Beyond differences in relative frequency of intrinsic subtypes, black and white patients had distinct gene expression, protein expression, and somatic mutation landscapes. However, the majority of these molecular differences were eliminated after adjusting for subtype; in the subtype-adjusted models, we found 142 genes, 16 methylation probes, 4 copy number segments, 1 protein, and no somatic mutation were differentially expressed or present between black and white patients. Using the top 40 differentially expressed genes, we built a race-enriched gene signature, which had excellent capacity of distinguishing breast tumors from black versus white patients (c-index=0.852 in the validation dataset). We also estimated the heritability of breast cancer subtype (basal vs. non-basal) to be 0.436 (p=1.5x10-14) and showed that two genetic variants (rs1078806 in FGFR2, rs34084277 in BABAM1) were associated with intrinsic subtype and can partially explain racial differences in subtype frequencies.

Conclusion: On the molecular level, once intrinsic subtype frequency differences are accounted for, there are few genomic or proteomic differences observed between blacks and whites. More than 40% of breast cancer subtype frequency differences may be due to genetic ancestry. These results suggest that future studies are warranted to investigate genetic and non-genetic factors that contribute to the development and progression of breast cancer subtypes in order to reduce racial disparity.

Citation Format: Huo D, Hu H, Rhie SK, Gamazon ER, Cherniack AD, Liu J, Yoshimatsu TF, Pitt JJ, Hoadley KA, Troester M, Ru Y, Lichtenberg T, Sturtz LA, Shelley CS, Mills GB, Laird PW, Shriver CD, Perou CM, Olopade OI. Comprehensive comparison of breast cancer molecular portraits by African and European ancestry in the cancer genome atlas [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P1-05-11.

Endometrial Carcinoma Recurrence Score (ECARS) validates to identify aggressive disease and associates with markers of epithelial-mesenchymal transition and PI3K alterations

PURPOSE

Our previously reported 29-gene expression signature identified an aggressive subgroup of endometrial cancer patients with PI3K activation. We here wanted to validate these findings by independent patient series.

PATIENTS AND METHODS

The 29-gene expression signature was assessed in fresh frozen tumor tissue from 280 primary endometrial carcinomas (three independent cohorts), 19 metastatic lesions and in 333 primary endometrial carcinomas using TCGA data, and expression was related to clinico-pathologic features and survival. The 29-gene signature was assessed by real-time quantitative PCR, DNA oligonucleotide microarrays, or RNA sequencing. PI3K alterations were assessed by immunohistochemistry, DNA microarrays, DNA sequencing, SNP arrays or fluorescence in situ hybridization. A panel of markers of epithelial-mesenchymal transition (EMT) was also correlated to the 29-gene signature score.

RESULTS

High 29-gene Endometrial Carcinoma Recurrence Score (ECARS) values consistently validated to identify patients with aggressive clinico-pathologic phenotype and reduced survival. Within the presumed favorable subgroups of low grade, endometrioid tumors confined to the uterus, high ECARS still predicted a poor prognosis. The score was higher in metastatic compared to primary lesions (P<0.001) and was significantly associated with potential measures of PI3K activation, markers of EMT and vascular invasion as an indicator of metastatic spread (all P<0.001).

CONCLUSIONS

ECARS validates to identify aggressive endometrial carcinomas in multiple, independent patients cohorts. The higher signature score in metastatic compared to primary lesions, and the potential link to PI3K activation and EMT, support further studies of ECARS in relation to response to PI3K and EMT inhibitors in clinical trials of metastatic endometrial carcinoma.

G(alpha)11 signaling through ARF6 regulates F-actin mobilization and GLUT4 glucose transporter translocation to the plasma membrane

The action of insulin to recruit the intracellular GLUT4 glucose transporter to the plasma membrane of 3T3-L1 adipocytes is mimicked by endothelin 1, which signals through trimeric G(alpha)q or G(alpha)11 proteins. Here we report that murine G(alpha)11 is most abundant in fat and that expression of the constitutively active form of G(alpha)11 [G(alpha)11(Q209L)] in 3T3-L1 adipocytes causes recruitment of GLUT4 to the plasma membrane and stimulation of 2-deoxyglucose uptake. In contrast to the action of insulin on GLUT4, the effects of endothelin 1 and G(alpha)11 were not inhibited by the phosphatidylinositol 3-kinase inhibitor wortmannin at 100 nM. Signaling by insulin, endothelin 1, or G(alpha)11(Q209L) also mobilized cortical F-actin in cultured adipocytes. Importantly, GLUT4 translocation caused by all three agents was blocked upon disassembly of F-actin by latrunculin B, suggesting that the F-actin polymerization caused by these agents may be required for their effects on GLUT4. Remarkably, expression of a dominant inhibitory form of the actin-regulatory GTPase ARF6 [ARF6(T27N)] in cultured adipocytes selectively inhibited both F-actin formation and GLUT4 translocation in response to endothelin 1 but not insulin. These data indicate that ARF6 is a required downstream element in endothelin 1 signaling through G(alpha)11 to regulate cortical actin and GLUT4 translocation in cultured adipocytes, while insulin action involves different signaling pathways.

Function of the Neurospora crassa mitochondrial tyrosyl-tRNA synthetase in RNA splicing. Role of the idiosyncratic N-terminal extension and different modes of interaction with different group I introns

The Neurospora crassa mitochondrial tyrosyl-tRNA synthetase (CYT-18 protein) promotes the splicing of group I introns by helping the intron RNA fold into the catalytically active structure. The regions required for splicing include an idiosyncratic N-terminal extension, the nucleotide-binding fold domain, and the C-terminal RNA-binding domain. Here, we show that the idiosyncratic N-terminal region is in fact comprised of two functionally distinct parts: an upstream region consisting predominantly of a predicted amphipathic alpha-helix (H0), which is absent from bacterial tyrosyl-tRNA synthetases (TyrRSs), and a downstream region, which contains predicted alpha-helices H1 and H2, corresponding to features in the X-ray crystal structure of the Bacillus stearothermophilus TyrRS. Bacterial genetic assays with libraries of CYT-18 mutants having random mutations in the N-terminal region identified functionally important amino acid residues and supported the predicted structures of the H0 and H1 alpha-helices. The function of N and C-terminal domains of CYT-18 was investigated by detailed biochemical analysis of deletion mutants. The results confirmed that the N-terminal extension is required only for splicing activity, but surprisingly, at least in the case of the N. crassa mitochondrial (mt) large ribosomal subunit (LSU) intron, it appears to act primarily by stabilizing the structure of another region that interacts directly with the intron RNA. The H1/H2 region is required for splicing activity and TyrRS activity with the N. crassa mt tRNA(Tyr), but not for TyrRS activity with Escherichia coli tRNA(Tyr), implying a somewhat different mode of recognition of the two tyrosyl-tRNAs. Finally, a CYT-18 mutant lacking the N-terminal H0 region is totally defective in binding or splicing the N. crassa ND1 intron, but retains substantial residual activity with the mt LSU intron, and conversely, a CYT-18 mutant lacking the C-terminal RNA-binding domain is totally defective in binding or splicing the mt LSU intron, but retains substantial residual activity with the ND1 intron. These findings lead to the surprising conclusion that CYT-18 promotes splicing via different sets of interactions with different group I introns. We suggest that these different modes of promoting splicing evolved from an initial interaction based on the recognition of conserved tRNA-like structural features of the group I intron catalytic core.

The protein-tyrosine kinase fer associates with signaling complexes containing insulin receptor substrate-1 and phosphatidylinositol 3-kinase

In a screen for 3T3-F442A adipocyte proteins that bind SH2 domains, we isolated a cDNA encoding Fer, a nonreceptor protein-tyrosine kinase of the Fes/Fps family that contains a functional SH2 domain. A truncated splicing variant, iFer, was also cloned. iFer is devoid of both the tyrosine kinase domain and a functional SH2 domain but displays a unique 42-residue C terminus and retains the ability to form oligomers with Fer. Expression of both Fer and iFer proteins are strikingly increased upon differentiation of 3T3-L1 fibroblasts to adipocytes. Platelet-derived growth factor treatment of the cultured adipocytes caused rapid tyrosine phosphorylation of Fer and its recruitment to complexes containing platelet-derived growth factor receptor and the p85 regulatory subunit of phosphatidylinositol (PI) 3-kinase. Insulin treatment of 3T3-L1 adipocytes stimulated association of Fer with complexes containing tyrosine phosphorylated IRS-1 and PI 3-kinase but did not stimulate tyrosine phosphorylation of Fer. PI 3-kinase activity in anti-Fer immunoprecipitates was also acutely activated by insulin treatment of cultured adipocytes. These data demonstrate the presence of Fer tyrosine kinase in insulin signaling complexes, suggesting a role of Fer in insulin action.

Role of the Raf/mitogen-activated protein kinase pathway in p21ras desensitization

Desensitization of p21(ras) after stimulation of cells by growth factors and phorbol 12-myristate 13-acetate (PMA) correlates with hyperphosphorylation of the guanine nucleotide exchange factor Son-of-sevenless (Sos) and its dissociation from the adaptor protein Grb2 (Cherniack, A., Klarlund, J. K., Conway, B. R., and Czech, M. P. (1995) J. Biol. Chem. 270, 1485-1488). To test the role of the Raf/mitogen-activated protein (MAP) kinase pathway, we utilized cells expressing a chimera composed of the catalytic domain of p74Raf-1 and the hormone binding domain of the estradiol receptor (DeltaRaf-1:ER). Estradiol markedly stimulated DeltaRaf-1:ER and the downstream MEK and MAP kinases in these cells as well as Sos phosphorylation. However, the dissociation of Grb2 from Sos observed in response to PMA was not apparent upon DeltaRaf-1:ER activation. Furthermore, stimulation of DeltaRaf-1:ER did not impair GTP loading of p21(ras) in response to platelet-derived growth factor or epidermal growth factor. We conclude that activation of the Raf/MAP kinase pathway alone in these cells is insufficient to cause disassembly of Sos from Grb2 or to interrupt the ability of Sos to catalyze activation of p21(ras).

Divergent mechanisms for homologous desensitization of p21ras by insulin and growth factors

Previous work suggested that desensitization of p21ras in response to growth factors such as epidermal growth factor (EGF) results from receptor down-regulation. Here we show that p21ras is desensitized by insulin in 3T3-L1 adipocytes in the continued presence of activated insulin receptors, while loss of epidermal growth factor and platelet-derived growth factor (PDGF) receptors in response to their ligands correlates with p21ras desensitization. Furthermore, elevated amounts of Grb2/Shc complexes persisted throughout p21ras desensitization by insulin. However, immunoblotting of anti-Son-of-sevenless (Sos) 1 and 2 immunoprecipitates with anti-Grb2 antisera revealed that p21ras desensitization in response to insulin and PDGF, but not EGF, is associated with a marked decrease in cellular complexes containing Sos and Grb2 proteins. Nonetheless, the desensitization of p21ras in response to these stimuli was homologous, in that each peptide could reactivate [32P]GTP loading of p21ras after desensitization by any of the others. Taken together, these data indicate that insulin, EGF, and PDGF all cause disassembly of Sos proteins from signaling complexes during p21ras desensitization, but at least two mechanisms are involved. Insulin elicits dissociation of Sos from Grb2 SH3 domains, whereas EGF signaling is reversed by receptor down-regulation and Shc dephosphorylation, releasing Grb2 SH2 domains. PDGF action triggers both mechanisms of Grb2 disassembly, which probably operate in concert with GAP to attenuate p21ras signaling.

Disassembly of Son-of-sevenless proteins from Grb2 during p21ras desensitization by insulin

Insulin receptor signaling acutely stimulates GTP loading of p21ras, apparently by mobilizing complexes of Grb2 and the guanine nucleotide exchangers Son-of-sevenless (Sos) 1 and 2 to associate with tyrosine-phosphorylated proteins in the plasma membrane. Here we show that in 32P-labeled 3T3-L1 adipocytes the elevated cellular concentrations of [32P]GTP-bound p21ras in response to insulin return to near basal levels after 20-30 min of hormone stimulation, while insulin receptors remain activated. Lysates of such desensitized cells were quantitatively immunoprecipitated with an antiserum recognizing both Sos1 and Sos2 proteins or a specific anti-Sos2 antiserum. Immunoblot analysis of these precipitates revealed that insulin causes a marked hyperphosphorylation of Sos1 and a 50% decrease in Grb2 associated with Sos proteins under these conditions. Similarly, anti-Grb2 immunoprecipitates of such lysates revealed the presence of decreased Sos1 protein due to insulin action. The disassembly of Grb2 from Sos proteins slightly precedes the time course of p21ras deactivation in response to insulin. These data are consistent with the hypothesis that the dissociation of Grb2 from Sos proteins caused by insulin in 3T3-L1 cells mediates p21ras deactivation and desensitization.

Phosphorylation of the Ras nucleotide exchange factor son of sevenless by mitogen-activated protein kinase

Son of sevenless-1 and -2 (Sos-1 and -2) are guanosine nucleotide exchange factors implicated in the activation of Ras by both the insulin and epidermal growth factor signal transduction pathways. Ras appears to function by initiating the activation of cellular protein kinases including mitogen-activated protein (MAP) kinases. Sos proteins contain numerous sequences in their carboxyl-terminal regions which correspond to consensus sites for MAP kinase phosphorylation. To examine whether these sites are substrates for MAP kinases, the cDNA encoding Drosophila Sos (dSos) was tagged with sequences encoding the major antigenic epitope of the influenza virus hemagglutinin (HA) to create a dSosHA fusion construct. dSosHA was transiently expressed in COS-1 cells and immunoprecipitated with anti-HA antibodies. When immune complexes were incubated with purified MAP kinase and [gamma-32P]ATP, a phosphorylated band of 180 kDa was observed when analyzed by SDS-polyacrylamide gel electrophoresis. This band was not present in immunoprecipitations from cells transfected with vector alone. No phosphorylation of the 180 kDa band was seen when immunoprecipitates were incubated with [gamma-32P]ATP in the absence of MAP kinase. Two dimensional analysis of tryptic peptides from dSosHA phosphorylated by MAP kinase in vitro revealed two major phosphorylated species that were also found in dSosHA isolated from COS-1 cells labeled with 32Pi. These results are consistent with the hypothesis that a feedback loop exists wherein growth factor-activated MAP kinases phosphorylate and regulate Sos proteins.

Binding of the Ras activator son of sevenless to insulin receptor substrate-1 signaling complexes

Signal transmission by insulin involves tyrosine phosphorylation of a major insulin receptor substrate (IRS-1) and exchange of Ras-bound guanosine diphosphate for guanosine triphosphate. Proteins containing Src homology 2 and 3 (SH2 and SH3) domains, such as the p85 regulatory subunit of phosphatidylinositol-3 kinase and growth factor receptor-bound protein 2 (GRB2), bind tyrosine phosphate sites on IRS-1 through their SH2 regions. Such complexes in COS cells were found to contain the heterologously expressed putative guanine nucleotide exchange factor encoded by the Drosophila son of sevenless gene (dSos). Thus, GRB2, p85, or other proteins with SH2-SH3 adapter sequences may link Sos proteins to IRS-1 signaling complexes as part of the mechanism by which insulin activates Ras.

The mitochondrial tyrosyl-tRNA synthetase of Podospora anserina is a bifunctional enzyme active in protein synthesis and RNA splicing

The Neurospora crassa mitochondrial tyrosyl-tRNA synthetase (mt tyrRS), which is encoded by the nuclear gene cyt-18, functions not only in aminoacylation but also in the splicing of group I introns. Here, we isolated the cognate Podospora anserina mt tyrRS gene, designated yts1, by using the N. crassa cyt-18 gene as a hybridization probe. DNA sequencing of the P. anserina gene revealed an open reading frame (ORF) of 641 amino acids which has significant similarity to other tyrRSs. The yts1 ORF is interrupted by two introns, one near its N terminus at the same position as the single intron in the cyt-18 gene and the other downstream in a region corresponding to the nucleotide-binding fold. The P. anserina yts1+ gene transformed the N. crassa cyt-18-2 mutant at a high frequency and rescued both the splicing and protein synthesis defects. Furthermore, the YTS1 protein synthesized in Escherichia coli was capable of splicing the N. crassa mt large rRNA intron in vitro. Together, these results indicate that YTS1 is a bifunctional protein active in both splicing and protein synthesis. The P. anserina YTS1 and N. crassa CYT-18 proteins share three blocks of amino acids that are not conserved in bacterial or yeast mt tyrRSs which do not function in splicing. One of these blocks corresponds to the idiosyncratic N-terminal domain shown previously to be required for splicing activity of the CYT-18 protein. The other two are located in the putative tRNA-binding domain toward the C terminus of the protein and also appear to be required for splicing. Since the E. coli and yeast mt tyrRSs do not function in splicing, the adaptation of the Neurospora and Podospora spp. mt tyrRSs to function in splicing most likely occurred after the divergence of their common ancestor from yeast.

Function of Neurospora mitochondrial tyrosyl-tRNA synthetase in RNA splicing requires an idiosyncratic domain not found in other synthetases

Neurospora mitochondrial tyrosyl-tRNA synthetase (mt TyrRS), which is encoded by nuclear gene cyt-18, functions in splicing group I introns. Analysis of intragenic partial revertants of the cyt-18-2 mutant and in vitro mutants of the cyt-18 protein expressed in E. coli showed that splicing activity of the cyt-18 protein is dependent on a small N-terminal domain that has no homolog in bacterial or yeast mt TyrRSs. This N-terminal splicing domain apparently acts together with other regions of the protein to promote splicing. Our findings support the hypothesis that idiosyncratic sequences in aminoacyl-tRNA synthetase may function in processes other than aminoacylation. Furthermore, they suggest that splicing activity of the Neurospora mt TyrRs was acquired after the divergence of Neurospora and yeast, and they demonstrate one mechanism whereby splicing factors may evolve from cellular RNA binding proteins.

Effects of verapamil on histamine-and carbachol-induced contraction of pulmonary tissues in vitro

It has been hypothesized that calcium antagonists may be useful in the management of airway hyperreactivity. In these studies, we evaluated the effects of verapamil on guinea pig tracheal spirals and parenchymal lung strips in vitro. Preincubation of both tissues with verapamil caused concentration-dependent inhibition of contraction with significant effects noted at a 10 micromolar concentration. At this concentration of verapamil, approximately fivefold greater concentrations of either histamine or carbachol were required to produce contraction of tracheal spirals; and 21-fold greater concentrations of histamine and 630-fold greater concentrations of carbachol were required to contract parenchymal strips. We also assessed the ability of verapamil to reverse contraction. Significant reversal of both histamine- and carbachol induced contraction was observed with concentrations of 3 micromolar verapamil and contraction was nearly abolished with a 100 micromolar concentration. These data demonstrate that verapamil can both inhibit airway contraction and reverse contraction once it is present and further suggest that verapamil or other calcium antagonists may prove useful in the management of airway hyperreactivity.

Clotrimazole, an inhibitor of epidermal benzo(a)pyrene metabolism and DNA binding and carcinogenicity of the hydrocarbon

Clotrimazole, a topically applied imidazole antifungal agent widely used in dermatological practice, was shown to be a potent inhibitor of the epidermal metabolism of benzo(a)pyrene (BP) and its microsomal enzyme-mediated binding both to neonatal rat epidermal DNA in vivo and to calf thymus DNA in vitro. Varying concentrations of clotrimazole added to in vitro incubation systems resulted in a dose-dependent inhibition of cytochrome P-450-dependent microsomal aryl hydrocarbon hydroxylase (AHH) in control animals as well as in animals pretreated with topical application of known inducers of the enzyme. Inhibition of epidermal AHH by topically applied clotrimazole was time and dose dependent. The 50% inhibition of clotrimazole for epidermal AHH ranged from 0.12 to 0.25 microM, which suggests that clotrimazole is among the most potent inhibitors of epidermal AHH yet identified. Clotrimazole was also found to be a potent inhibitor of epoxide hydrolase activity in vitro with a 50% inhibition at 0.1 mM. High-pressure liquid chromatographic analysis of the metabolism of BP in rat epidermal microsomes revealed substantial inhibition of metabolite formation by clotrimazole. This occurred in microsomes prepared from untreated as well as animals pretreated with inducers of the enzyme. Furthermore, a single topical application of clotrimazole resulted in 80 and 30% induction of epidermal and hepatic glutathione S-transferase activity, respectively. Topical application of clotrimazole to the skin of BALB/c mice substantially increased the latent period for the development of skin tumors by 3-methylcholanthrene. These studies indicate that clotrimazole is an extremely potent inhibitor of epidermal BP metabolism and of the DNA-binding of polycyclic aromatic hydrocarbon (PAH) carcinogens, and is an enhancer of enzymes necessary for detoxification of the PAH. Clotrimazole also reduces the formation of carcinogenic and mutagenic metabolites of BP in vitro and in vivo and inhibits induction of skin tumors by the PAH. These data indicate that the imidazole antifungal clotrimazole offers promise as an agent useful for the modulation of PAH cancer risk in the skin.