*3.5. Prognostic Significance of HIP1R and PD-L1*

Patients with high HIP1R expression had inferior PFS to patients with low HIP1R expression (*p* = 0.037, Figure 5A). Patients with high HIP1R expression also showed an inferior OS than patients with low HIP1R expression, however the statistical significance was not reached (*p* = 0.11, Figure 5B). Patients with high PD-L1 expression had superior PFS or OS than patients with low PD-L1 expression (*p* = 0.028, Figure 5C and *p* = 0.031, Figure 5D, respectively). Furthermore, patients with high HIP1R expression and low PD-L1 expression had lower PFS or OS than patients with other expression patterns (*p* < 0.001, Figure 5E and *p* = 0.001, Figure 5F, respectively). In multivariate analysis, high HIP1R expression was an independent prognostic factor for PFS (HR = 2.098, *p* = 0.019, Table 3).

**Figure 5.** Comparison of survival rates according to HIP1R and PD-L1 expression. (**A**) Progression-free survival (PFS) and expression of HIP1R. (**B**) Overall survival (OS) and expression of HIP1R (**C**) PFS and PD-L1. (**D**) OS and PD-L1. (**E**) PFS, HIP1R, and PD-L1. (**F**) OS, HIP1R, and PD-L1.


**Table 3.** Univariate and multivariate analyses of progression-free survival.

† Cox proportional hazards regression model. Abbreviations: CI, confidence interval; EGFR, epidermal growth factor receptor; HIP1R, Huntingtin Interacting Protein 1 Related; HR, hazard ratio; PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1.

We used a Kaplan Meier plotter tool and performed survival analysis according to HIP1R mRNA expression in lung adenocarcinoma and lung squamous cell carcinoma patients. The group with high HIP1R mRNA expression exhibited poorer OS in patients with adenocarcinoma (*p* = 0.026, Figure S4A). However, in lung squamous cell carcinoma, HIP1R mRNA expression was not correlated with OS (*p* = 0.63, Figure S4B)

### **4. Discussion**

This study had several novel discoveries. First, we found that the expression of HIP1R was an independent predictive factor for anti-PD-1 treatment response by NSCLC patients. Second, the expression of HIP1R was an independent prognostic factor of PFS in patients treated with anti-PD-1 inhibitors. Third, GSEA revealed that *HIP1R* mRNA expression was tightly correlated with immune-related gene sets in lung adenocarcinoma. These GSEA results suggested that *HIP1R* mRNA expression plays an important role in regulating the expression of PD-L1.

GSEA revealed that low *HIP1R* mRNA expression was closely associated with allograft rejection, inflammatory responses, IL6-JAK-STAT3, IL2-STAT5, and interferon gamma response pathways in lung adenocarcinoma. PD-L1 expression is correlated with marked expression of adaptive immune responses (CD8+ T-cells) [21]. In our study, CD8 was also included in the core enrichment gene list of HALLMARK\_ALLOGRAFT\_REJECTION. Previous studies have also reported that the IL6-JAK-STAT3 pathway induces PD-L1 upregulation. IL-6 is positively correlated with PD-L1 expression in human hepatocellular carcinoma (HCC) cells, and IL-6 induces PD-L1 stability through glycosylation in a HCC cell line [22]. Glioblastoma-derived IL6 is required for up-regulation of myeloid PD-L1 in glioblastoma through a STAT3-dependent mechanism [23]. Combined blockade of IL6 and PD-L1 signaling achieves synergistic antitumor immune responses in colon carcinoma and murine melanoma models [24]. PD-L1 expression is also regulated by interferon gamma signaling in a melanoma cell line [25]. GSEA suggested that HIP1R expression plays an important role in adaptive immune responses associated with PD-L1.

In the present study, no correlation was identified between HIP1R and PD-L1 protein expression, However, HIP1R mRNA expression was negatively correlated with the mRNA expression level of PD-L1 in adenocarcinoma and squamous cell carcinoma. There are several possible explanations for this discrepancy. Post-transcriptional and post-translational modifications affect the level of protein expression [26]. Proteins can have significantly different half-lives in vivo [27]. There were no cases of surgery in our study, therefore the only sample we received was a biopsy. We cannot conduct additional experiments for mRNA testing of HIP1R and PD-L1, because very little tumor tissue remains in paraffin tissue.

Patients with high HIP1R mRNA expression exhibited poor clinical outcomes in web-based mRNA data of adenocarcinoma cases; however, HIP1R mRNA expression was not correlated with OS in squamous cell carcinoma. From our IHC data, HIP1R expression was correlated with poor clinical outcomes. However, we did perform subgroup analysis according to histologic type because of our small sample size. HIP1R levels also correlate with a set of genes that reflect PD-L1-related immune pathways in GSEA analysis of adenocarcinoma cases; however, there were no statistically significant immune-related gene sets associated with HIP1R mRNA expression in squamous cell carcinoma. Currently, lung adenocarcinoma and squamous cell carcinoma are known to involve different biologic mechanisms and prognoses. Therefore, the role of HIP1R in adaptive immune responses and its effect on clinical outcomes may vary depending on the histological type.

Despite some surprising discoveries, our study has certain limitations. First, our cohort was small. We performed multivariate logistic regression and prognostic analyses on only 52 samples. However, the web-based mRNA dataset also revealed results similar to ours. These results encourage further investigations involving larger populations. Second, we used an IHC method to detect HIP1R protein expression. There is no information regarding standardization, reliability, and reproducibility of IHC staining. We used the same antibody that Wang et al. used [12]. However, Wang et al. used HIP1R antibody (16814-1-AP) in Western Blott (WB) and immunofluorescence alone. Only recently has HIP1R attracted attention in cancer research, so few studies have been done on HIP1R. Therefore, there are no antibodies that are commonly used in immunohistochemistry. In the catalog of HIP1R antibody (16814-1-AP), it can be used in IHC, immunoprecipitation (IP), WB, and ELISA. According to the manufacturer's guidelines, this antibody was validated by western blot in HeLa cells and human liver tissue. We used a human placenta tissue as positive control as recommended. An automatic IHC staining

device (Benchmark XT) may improve the reproducibility of IHC staining. The H-scoring method is widely used for immunochemical staining, and is known to have relatively high reproducibility among pathologists [28,29]. Third, we examined the protein expression of HIP1R and PD-L1 in refractory advanced NSCLC; however, the mRNA profiles of HIP1R and PD-L1 were not evaluated. Because protein expression of HIP1R did not correlate with PD-L1 expression, the relationship between HIP1R and PD-L1 mRNA expression profile is very important. To verify the results of GSEA, we should evaluate the mRNA expression profiles of HIP1R. However, the sample we have is a small biopsy, and we have already performed several immunohistochemical stainings for diagnosis and ALK and EGFR mutation tests. Therefore, currently, very little tumor tissue remains in paraffin tissue and we cannot conduct additional experiments for mRNA testing. To confirm our experiments, future research should measure the mRNA expression level of HIP1R on many samples and investigate the relationship with the PD-1 blocker and PD-L1 expression.

In conclusion, we examined the expression of HIP1R in 52 refractory NSCLC samples from patients treated with PD-1 inhibitors. HIP1R expression was an independent biomarker predicting patient response to PD-1 inhibitors. High HIP1R expression was an independent predictor of poor PFS. In addition, HIP1R mRNA expression was significantly correlated with immune-related gene sets in lung adenocarcinoma. These immune-related gene sets are known to play important roles in PD-L1 regulation. Based on our findings, HIP1R expression may be a promising predictor for the therapeutic determination of responses to anti-PD-1 treatment.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2077-0383/9/5/1425/s1, Table S1. Gene sets within the top 20-ranked list related to low HIP1R in adenocarcinoma patients; Table S2. Core enrichment gene list of HALLMARK\_ALLOGRAFT\_REJECTION; Table S3. Core enrichment gene list of HALLMARK \_INFLAMMATORY\_RESPONSE; Table S4. Core enrichment gene list of HALLMARK\_IL6\_JAK\_STAT3\_SIGNALING; Table S5. Core enrichment gene list of HALLMARK\_IL2\_STAT5\_SIGNALING; Table S6. Gene sets within the top 20-ranked list related to low HIP1R in squamous cell carcinoma patients; Figure S1. HIP1R expression in positive control. Positive HIP1R expression in placental tissue, X400; Figure S2. PD-L1 sp263 expression in nonsmall cell carcinoma. (A) No staining of PD-L1 sp263, x400. (B) Faint PD-L1 sp263 staining, x400. (C) Moderate PD-L1 sp263 staining, x400. (D) Strong PD-L1 sp263 staining, x400; Figure S3. PD-L1 22c3 expression in nonsmall cell carcinoma. (A) No staining of PD-L1 22c3, x400. (B) Faint PD-L1 22c3 staining, x400. (C) Moderate PD-L1 22c3 staining, x400. (D) Strong PD-L1 22c3 staining, x400; Figure S4. Comparison of survival rates, according to HIP1R mRNA expression in patients with nonsmall cell carcinoma. (A) Overall survival (OS) and HIP1R in lung adenocarcinoma. (B) OS and HIP1R in lung squamous cell carcinoma.

**Author Contributions:** Conceptualization, Y.W.K.; formal analysis, Y.W.K.; funding acquisition, Y.W.K.; investigation, Y.W.K., J.-H.H., S.H., and H.W.L.; methodology, Y.W.K., J.-H.H., S.H., and H.W.L.; resources, Y.W.K.; supervision, Y.W.K.; visualization, Y.W.K.; writing—original draft, Y.W.K., J.-H.H., S.H., and H.W.L.; writing—review and editing, Y.W.K., J.-H.H, S.H., and H.W.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by Basic Science Research Program through the National Research Foundation of Korea(NRF) funded by the Ministry of Science, ICT (NRF-2017R1C1B5076342 for Young Wha Koh) and the faculty research fund (Ajou translational research fund 2018) of Ajou University School of Medicine to Young Wha Koh and Seokjin Haam (M-2018-C0460-00035).

**Conflicts of Interest:** The authors declare no conflict of interest.
