Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2025 Feb 14;15:5469. doi: 10.1038/s41598-025-88972-8

Using major histocompatibility complex (MHC) II expression to predict antitumor response to CD4 + lymphocyte depletion

Yanping Wang 1, Minhyung Kim 1, Shengchen Su 1, Lahiruni Halwatura 1, Sungyong You 1, Hyung L Kim 1,✉
PMCID: PMC11828847  PMID: 39953106

Abstract

Immunotherapy targeting CD4 + FoxP3 + regulatory T cells (Tregs) through CD4 + lymphocyte depletion is being investigated as cancer treatment. This study examines the efficacy and limitations of CD4 depletion across various cancer models. We examined CD4 depletion in syngeneic mouse tumor models including B16 melanoma, Renca kidney cancer, and Hepa1-6 hepatocellular carcinoma. The study explored the relationship between MHC II expression and tumor growth, immune response, and survival. In mouse models, CD4 depletion suppressed B16 and Renca tumor growth but accelerated Hepa1-6 tumors. Hepa1-6 had high MHC II expression while the other two lines had low MHC II expression. Manipulation of MHC II expression in Hepa1-6 and Renca cell lines confirmed the mechanistic importance of MHC II in generating a CD4-based antitumor immune response. Analysis of the TCGA dataset supports the relevance of this mechanism in patients. In tumors with high MHC II and low MHC I expression, increased CD4 levels correlated with improved survival. CD4 depletion can suppress tumor growth or promote tumor growth, depending on tumor MHC expression status. MHC status may identify tumors where intratumor CD4 may be a better early-response marker than CD8.

Keywords: CD4 depletion, MHC II, Tumor immunology

Subject terms: Tumour immunology, Tumour biomarkers

Introduction

Immunotherapy is becoming a cornerstone for cancer treatment. Among promising approaches under investigation is the depletion of CD4 + lymphocytes as a strategy to target CD4 + FoxP3 + regulatory T cells (Tregs).1–5 Tregs are a pivotal regulator of the immune response and functions as a critical immune checkpoint. However, despite their significance, there is currently no FDA-approved cancer therapy that directly and specifically targets Tregs. Therefore, CD4 depletion has been proposed as a strategy to control Tregs.1–5 This strategy is based on the premise that by the time a malignancy is diagnosed, the immune system has been primed by CD4 + effector T cells and the vast majority of CD4 + T cells are immunosuppressive. Therefore, CD4 + T cells are no longer needed for antitumor immunity. In previously reported preclinical models and early phase clinical testing, CD4 depletion is effective in stimulating antitumor immunity.1–5.

However, the interplay between tumor and host immunity is heterogeneous, and CD4_ lymphocytes have been reported to control tumor growth in some settings.6,7 This study explored the efficacy and limitations of CD4 depletion in controlling tumor growth across various syngeneic cancer models. We observed that CD4 depletion suppressed the growth of B16 melanomas and Renca kidney cancers. In contrast, it accelerated the growth of Hepa1-6 hepatocellular carcinomas. Amongst our tumor models, Hepa1-6 had uniquely high MHC II expression, suggesting that MHC II determines how CD4 depletion affects tumor growth. We confirm the mechanistic relationship between MHC II expression and CD4 + cells by manipulating MHC II expressions in Hepa1-6 and Renca cell lines and noting in vitro and in vivo changes in tumor growth and immune-mediated tumor killing.

Our observations in mouse tumor models indicate that MHC I and MHC II expression may identify human tumors where tumor growth suppression depends on CD4 + or CD8 + lymphocytes. This hypothesis was supported by our analysis of survival outcomes from the TCGA pancancer dataset. Specifically, in tumors with high MHC I expression, higher CD8 expression predicted better survival and higher CD4 levels predicted worse survival. However, as expected, in tumors with low MHC I expression and high MHC II expression, higher tumor CD4 levels predicted survival in the opposite direction, predicting improved survival.

Methods

Tumor cell and mice

Renca, B16F10, Hepa1-6 cancer cell lines were purchased from ATCC (Manassas, Virginia). The cells were maintained in RPMI 1640 or DMEM medium supplemented with 10% heat-inactivated FBS (Gemcell, West Sacramento, CA), 2 mmol/L of L-glutamine, 100 units/mL of penicillin, 100 μg/mL of streptomycin, and 250 ng/mL of Amphotericin B (ThermoFisher, Waltham, MA). These cells were periodically authenticated by morphologic and histologic inspection and animal grafting to assess their ability to grow and metastasize. The cells were annually tested for mycoplasma using Myco Alert Kit (Lonza, Allendale NJ).

Balb/C or C57BL/6 mice, 5–8 weeks old, were purchased from Jackson laboratory (Ellsworth, Maine) and housed under pathogen free conditions. All experiments involving animals followed federal and state standards in the federal Animal Welfare Act and the NIH guide for the care and use of laboratory animals. The IACUC approved this study (IACUC010003) for all animal experiments. The Renca tumors were generated by subcutaneously injecting 2 × 105 tumor cells into flanks of Balb/C mice. Hepa1-6 tumors were generated by subcutaneously injecting 2 × 106 tumor cells and B16 tumor were generated by subcutaneously injecting 2 × 105 tumor cells into flanks of B6 mice. Mouse CD8 cells were depleted with 200 μg αCD8 administered i.p. after injection of tumor cells, CD4 depletion was performed by i.p. injection of 200 μg αCD4 after tumor cells were injected. No mouse anesthesia was required, and euthanasia was performed by placing mice in a CO2 chamber (fill rate for 30–70% displacement of the chamber volume per minute with CO2) followed by cervical dislocation.

For all animal studies, 5 or more mice were used per group. Mice were genetically and phenotypically identical, they were randomly assigned to experimental groups, and no mice were excluded from analysis. After randomization, to minimize risk of treatment and measurement mistakes, investigators were not blinded to treatment allocation. Tumor volume was calculated using the following formula: V = W2xL/2, and no tumor measurements were excluded for data analysis. ARRIVE reporting guidelines were followed.8

Antibodies and reagents

The following monoclonal antibodies (mAb) with a fluorescent conjugate for flow cytometry were obtained from Biolegend (San Diego, CA): αMCH I H-2 kb (AF6-88.5), H-2Kd (SF1-1.1), αMHCII I-A/I-E (M5/114.15.2) αCD4 (GK 1.5 and RM4-5), αCD8 (53–6.7), αCD45 (30-F11), αIFN-γ (XMG1.2), αTNFα (TN3-19.12), αCD107 (ID4B). αCD8 (2.43) and αCD4 (GK 1.5) antibodies for T cell depletion were purchased from BioXcell (West Lebanon, NH). αMHCII (PA5-116,876) and goat anti-rabbit (A16101) antibody for immunofluorescence staining were purchased from Life Technologies (Carlsbad, CA). Brefeldin A was purchased from eBioscience / ThermoFisher Scientific (San Diego, CA), PMA was purchased from Selleck Chemicals LLC (Houston, TX). The following tumor dissociation reagents were purchased from Sigma-Aldrich (Saint Louis, MO): Collagenase type IV (Sigma, C5138), DNase type IV (20 U/mL, Sigma, D5025), and Hyaluronidase type V (0.1 mg/mL, Sigma, H6254).

CD4, CD8 isolation and adoptive transfer

To generate antitumor immunity, mice were treated with 2 doses of dendritic cell (DC) vaccine (generated with cell lysate) over 14 days. Spleen and lymph nodes were harvested from these mice, and lymphocytes were isolated using EasySep mouse CD4 or CD8 T cell isolation kit (STEMCELL technologies) and cultured in vitro with tumor lysate pulsed DC and IL-2 for 3 days. 2 × 106 activated CD4 or CD8 + cells were adoptively transferred by tail vein injection into recipient mice (5 mice per group), which were challenged with s.c. tumor cells. Tumor growth was measured every 2–3 days with caliper.

To generate the DC vaccine, 10 million Renca or Hepa1-6 cells were lysed using three freeze and thaw cycles and then centrifuged at 10,000 rpm; 1 mL of the supernatant was added to 10 million fresh DCs with LPS (1 μg/mL) for 24 h. DCs were harvested and washed 3 times with PBS and then used to stimulate lymphocytes.

CIITA was knocked down (KD) in Hepa 1–6 cells and over-expressed (OE) in Renca cells

The CIITA KD cells were created using MISSION® shRNA lentiviral transduction particles (Sigma-Aldrich, SHCLNV_NM_007575), following the manufacturer’s instructions. Five predesigned shRNA clones (TRCN0000086448, TRCN0000086449, TRCN0000086450, TRCN0000086451, and TRCN0000086452) in vector pLKO1 were purchased from Sigma-Aldrich. Hepa1-6 cells were transduced in the presence of 5 μg/mL of polybrene and selected using 500 μg/mL G418 to create stable lines.

CIITA OE cells were generated by transducing CIITA (NM_007575.3) mouse Tagged ORF Clone lentiviral particle (GeneCopoeia, LPP-Mm01492-Lv158-100) into Renca cells. The CIITA tagged ORF was inserted into pReceiver-Lv158-Neomycin vector. The Renca cells were transduced in the presence of 5 μg/mL polybrene and selected using 3 μg/mL puromycin to create stable lines.

T cells activation and T cell cytotoxicity

To prepare DC vaccine, mouse bone marrow was harvested from C57BL/6 mice and suspended in RPMI supplemented with 10% FBS. GM-CSF (10 ng/mL) was added to the medium and placed at 37 °C with 5% CO2. RPMI medium was changed every other day. DCs were harvested on day 7 and pulsed with tumor cell lysate for 16 h and activated with 10 μg/mL CpG for 4 h.

To assess mouse T cell activation following in vivo treatment, lymph nodes and spleen were collected to generate single cell suspensions and activated ex vivo with DC vaccine (fresh DC pulsed with tumor lysate) for 4 days. CD4, CD8, IFNγ were analyzed by flow cytometry (LSR II, BD Biosciences, San Jose, CA 95,131).

To assess T cell cytotoxicity in vitro, lymphocytes were harvested from WT Hepa1-6 tumor-bearing mice and co-cultured with CIITA WT or KD Hepa1-6 cells at various target: effector ratios for 18 h. Hepa1-6 cell death was quantified by flow cytometry by gating on CD45 negative cells and determining the percent of cells taking up propidium iodine.

Tumor cell dissociation and flow cytometry

Mouse tumors were cut into small pieces using surgical scissors and placed in 2.5 ml RPMI with collagenase type IV (1 mg/mL, Sigma, C5138), DNase type IV (20 U/mL, Sigma, D5025), and hyaluronidase type V (0.1 mg/mL, Sigma, H6254). The samples were processed using a gentleMACS™ dissociator with heaters (Miltenyi Biotec), according to the manufacturer’s protocol. Tumor-dissociated cells were harvested by centrifugation at 300 × g for 7 min and stained with αCD45 and αCD8 (BioLegend). Flow cytometry data were collected using the Symphony Flow Cytometer (BD Biosciences) and were analyzed using FlowJo v10 (BD Biosciences).

Immunofluorescence staining

Hepa 1–6 (WT and CIITA KD) or Renca (WT and CIITA OE) cells were grown on cover slides. When cell confluency reached 80%, the cells were fixed with 4% paraformaldehyde. After blocking with 5% bovine serum albumin (BSA), the slides were stained with a rabbit polyclonal anti-MHC II antibody (Life Technologies, PA5-116,876) and visualized with the goat anti-rabbit IgG (H + L) TRITC (Life Technologies, A16101). Mounting Medium with DAPI (Vector Laboratories, H-1200) was added to all slides before image capture using a Keyence BZ-X710 microscope.

TCGA Pan-cancer data analysis

RNA-sequencing data and corresponding clinical information of TCGA Pan-cancer samples were downloaded from University of California Santa Cruz Xena Browser9 using R package UCSCXenaTools10. The gene set z-score method11 was used to summarize the expression of genes belonging to MHC I (HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, HLA-G, and HLA-H) and MHC II (HLA-DMA, HLA-DMB, HLA-DOA, HLA-DOB, HLA-DPA1, HLA-DPB1, HLA-DQA1, HLA-DQA2, HLA-DQB1, HLA-DRA, HLA-DRB1, HLA-DRB3, HLA-DRB4, and HLA-DRB5). CD4 and CD8 T-cell score were estimated using EPIC method in R package immunedeconv (v2.1.0)12,13.

Statistical analysis

The Kaplan–Meier method was used to present survival functions. The Cox proportional hazard regression was used to estimate survival association. Hazard ratios (HRs) and 95% confidence intervals (CIs) were reported, and statistical significance was assessed based on nonoverlapping 95% CIs and the log-rank test. Statistical analyses were performed using R statistical software (v4.0.2; R Foundation, Vienna, Austria) and MATLAB (Mathworks, Natick, MA, USA) with two-sided tests and a significance level of 0.05. Error bars represent standard error of the mean.

Results

CD4 depletion has varied effects on tumor growth among tumor models

CD4 depletion, after tumor growth has already primed the immune system, has been shown to remove Tregs, stimulate antitumor immunity, and suppressive tumor growth in mouse models.1 However, we noted that this does not happen in all tumor models (Fig. 1, Supplemental Fig. 1). Tumors were established by subcutaneous injection of syngeneic tumor cells into mice, and after immune priming was expected to have occurred, CD4 depleting antibodies were administered and tumor growth was monitored (Fig. 1A). CD4 depletion suppressed the growth of B16 melanomas and Renca kidney cancers (Fig. 1B-C). However, Hepa1-6 hepatoma tumors, which normally do not grow in immunocompetent B6 mice, grew after CD4 depletion (Fig. 1D), suggesting CD4 cells suppress the growth of Hepa1-6 tumors. Our Hepa1-6 cells were newly purchased from ATCC, and injection of even 1 × 107 cells subcutaneously did not produce palpable tumor, which is consistent with a report by Hiotis et al14, where only intrahepatic injection resulted in tumor formation.

Fig. 1.

Fig. 1

Open in a new tab

Effect of CD4 depletion across tumor models. (A) Mice were injected s.c. with syngeneic tumor cells, and after Renca and Hepa1-6 tumors were palpable, mice were treated with 2 doses of CD4 depleting antibody, 200µg/mouse (typically on days 9 and 11), and tumor growth was monitored. (B) B16 melanoma cells were injected into B6 mice. (C) Renca kidney cancer cells were injected into Balb/c mice. (D) Hepa1-6 hepatocellular cells were injected into B6 mice. Statistical significance determined by repeated measures ANOVA. Supplemental Fig. 1 shows a duplicate experiment with each tumor growth shown as a separate line.

Suppression of Hepa1-6 tumor growth is dependent on CD4 + lymphocytes

To assess the role of CD4 + lymphocytes and CD8 + lymphocytes on Hepa1-6 tumor growth, CD4 + cells, CD8 + cells, or both cells were depleted after tumor cell injection in mice (Fig. 2A). As expected, CD4 depletion increased tumor growth while CD8 depletion did not significantly affect tumor growth (Fig. 2B). Interestingly, depleting both CD4 + and CD8 + cells caused the tumor to grow the fastest, suggesting that in the presence of CD4 depletion, CD8 + cells may assume a more important antitumor role.

Fig. 2.

Fig. 2

Open in a new tab

Role of CD4 + and CD8 + lymphocytes in Hepa1-6 tumor growth. (A) Depleting antibodies against CD4, CD8, or both (200µg of each antibody/mouse) were administered 1 day after s.c. injection of Hepa 1–6 cells. (B) Tumor growth was monitored. (C) After Hepa1-6 tumor were injected s.c., both CD4 and CD8 lymphocytes were depleted with antibodies as described for Figure A (recipient mice). From separate, non-tumor bearing mice, antitumor immunity was generated by treating with dendritic cells pulsed with Hepa1-6 lysate (donor mice). CD4 or CD8 cells were enriched from donor mouse splenocytes and lymphocytes. CD4, CD8, or both cells were adoptive transferred into tumor-bearing recipient mice on day 2. (D) Tumor growth was monitored. Statistical significance determined by repeated measures ANOVA. Representative results from duplicate experiments are shown.

To confirm the importance of CD4 + cells, both lymphocyte populations were depleted and then replaced (Fig. 2C). Adoptively transferred lymphocytes were harvested from nontumor-bearing (donor) mice stimulated with Hepa1-6 tumor lysate. After depleting both CD4 + cell and CD8 + cells in the tumor bearing (recipient) mice, replacing CD4 + cells or both CD4 + /CD8 + cells restored tumor growth-suppression (Fig. 2D). However, replacing CD8 + cells did not, confirming the importance of CD4 + cells in suppressing Hepa1-6 growth.

Hepa1-6 has high MHC II expression

Since CD4 functions as a bridge between the T cell receptor and MHC II, which can be found on tumor cells15–17 and antigen presenting cells, we assessed MHC I and MHC II expression in tumor cells. We hypothesized that MHC II expression on Hepa1-6 may allow CD4 + lymphocytes to suppress tumor growth. Consistent with this hypothesis, MHC II was highly expressed in only Hepa1-6 cells (Fig. 3A-B) , which is consistent with previous reports of MHC II expression in Hepa1-6 cell lines and tumors18. MHC II, but not MHC I, expression increased in Hepa1-6 cells in the presence of immune activating cytokines, INFγ and TNFα (Supplemental Fig. 2). With respect to the various tumor models we examined, the effect of CD4 depletion on tumor growth correlated with MHC II expression but not with MHC I expression.

Fig. 3.

Fig. 3

Open in a new tab

Tumor MHC I and II levels were assessed in mouse models and humans. (A) Hep1-6, B16 and Renca cells were assessed for MHC I expression by flow cytometry. (B) The same cell lines were assessed for MHC II expression. (C) Expression levels for MHC I and (D) MHC II of various cancer types using the TCGA pan-cancer dataset. Gene expressions were standardized using the z-score method. The box plot depicts the median as well as the upper and lower quartiles, and the whiskers depict 1.5 times the interquartile range.

These results also suggest that MHC expressions may be useful as clinical biomarkers. However, any candidate biomarker should have heterogenous expression to have potential to reflect biological heterogeneity. Interrogation of the TCGA pan-cancer dataset indicates that both MHC I and MHC II have heterogeneous expression both within and between tumor types (Fig. 3C).

MHC II expression was manipulated in tumor cell lines

Given these correlative observations, we sought to firmly establish MHC II expression as the mechanism that can produce differential responses to CD4 depletion. Therefore, we created cell lines where MHC II expression was either increased or decreased by manipulating CIITA, which is a transcriptional coactivator of MHCII (Fig. 4A). We confirmed in the TCGA pan-cancer dataset that CIITA and MHC II expressions are highly correlated (Fig. 4B). Hepa1-6 cells have high MHC II expression and shRNA knock down (KD) of CIITA reduced MHC II expression (Fig. 4C, Supplemental Fig. 3). Renca on the other hand has no MHC II expression and CIITA overexpression (OE) resulted in high MHC II expression (Fig. 4D, Supplemental Fig. 4). Both manipulated cell lines were characterized in our mouse models, which were subjected to CD4 depletion. We predicted that CIITA manipulation would change whether CD4 depletion would suppress or stimulate tumor growth.

Fig. 4.

Fig. 4

Open in a new tab

CIITA expression was knocked down and overexpressed. (A) CIITA is a transcriptional coactivator for MHC II. (B) MHC II expression and CIITA expression are highly correlated in the TCGA pan-cancer dataset with Spearman’s Rho of 0.8395 and p < 0.0001. (C) Wild type Hepa1-6 cells have high MHC II expression, and shRNA knock down (KD) reduced MHC II expression as assessed by immunocytochemistry and flow cytometry. (D) Wild type Renca cells have no MHC II expression, and transfection with a CIITA overexpression plasmid resulted in high MHC II expression as assessed by immunocytochemistry and flow cytometry.

MHC II KD in Hepa1-6 removed their dependence on CD4 + lymphocytes for tumor growth suppression

Prior to evaluating the growth of KD Hepa1-6 in mice, it was essential to confirm that genetic alteration did not impact their intrinsic growth rate. Therefore, the in vitro proliferation of WT and KD Hepa1-6 cells was compared and was not significantly different (Fig. 5A). However, when KD Hepa1-6 cells were injected into mice, they grew rapidly while the WT cells did not, suggesting that CD4 + lymphocytes could no longer control the KD Hepa1-6 tumors (Fig. 5B). To confirm that the KD cells were less capable of generating lymphocyte-mediated immunity, a killing assay was performed (Fig. 5C). In mice, the WT tumor cells generated a robust lymphocyte-mediated tumor-specific killing response while the KD tumor cells did not. Finally, we predicted that the KD Hepa1-6 tumors with low MHC II expression would be sensitive to CD4 depletion, like B16 and Renca tumors. This was in fact the case; while the WT Hepa1-6 tumors grew faster after CD4 depletion (Fig. 5D), the KD Hepa1-6 tumors were suppressed by CD4 depletion (Fig. 5E).

Fig. 5.

Fig. 5

Open in a new tab

Characterization of Hepa1-6 with CIITA KD. (A) In vitro growth rate was compared for CIITA knock down (KD) and WT Hepa1-6 cells. (B) KD and WT Hepa1-6 cells were injected s.c. into B6 mice and tumor growth was assessed. Representative results from duplicate experiments are shown. (C) For an immune-mediated tumor killing assay, lymphocytes were harvested from WT Hepa1-6 tumor-bearing mice, which were co-cultured with WT or KD Hepa1-6 cells. Hepa1-6 cell death was detected by flow cytometry by gating on CD45 negative cells and counting propidium iodine (PI) positive cells. (D) WT and (E) KD cells were injected s.c. into B6 mice to establish tumor, and tumor growth with and without CD4 lymphocyte depletion was compared. CD4 depleting antibodies were administered on days 9 and 11.

MHC II OE in Renca made them dependent on CD4 + lymphocytes for tumor growth suppression

The inverse experiments were performed with OE Renca cells. In vitro proliferation of WT and OE Renca cells were not significantly different (Fig. 6A). However, as expected, the in vivo growth of OE Renca was significantly decreased when compared to WT Renca (Fig. 6B). To better characterize the resulting immune response, splenocytes from these mice were compared. When compared to WT, OE Renca had significantly increased CD4 + cell and CD8 + splenocytes (Fig. 6C, Supplemental Fig. 5), although this may not necessarily reflect an increase in tumor-specific lymphocytes. Therefore, tumor-specific activation of CD4 + and CD8 + cells was assessed by measuring IFNγ expression following ex vivo restimulating with tumor-lysate pulsed DCs (Fig. 6D). In addition, immune activation should result in lymphocyte infiltration into tumor, and both CD4 + and CD8 + lymphocyte infiltration was increased in OE Renca tumors when compared to WT tumors (Fig. 6E, Supplemental Fig. 6).

Fig. 6.

Fig. 6

Open in a new tab

Characterization of Renca with CIITA overexpression. (A) In vitro growth rate was compared for CIITA overexpressing (OE) and WT Renca cells. (B) OE and WT Renca cells were injected s.c. into Balb/c mice and tumor growth was assessed. Representative results from duplicate experiments are shown. (C) Splenocytes were harvested from tumor-bearing mice. CD4 + and CD8 + cells were quantified as a percent of all splenocytes. (D) To assess CD8 + cell activation, CD4 + IFNγ + and CD8 + IFNγ + cells were quantified as a percent of splenocytes, after ex vivo stimulation with fresh DCs pulsed with tumor lysate. (E) CD45 + CD4 + and CD45 + CD8 + cells were quantified from tumor-dissociated cells. (F) WT Renca cells were injected s.c. into Balb/C mice to establish tumor, and tumor growth with and without CD4 lymphocyte depletion was compared. CD4 depleting antibodies were administered on days 9 and 11. (G) From these mice, splenocytes were harvested and CD4 + FoxP3 splenocytes were quantified. CD8 + IFNγ + splenocytes were quantified as described for Fig. 6D. (H–I) The experiments in Figs. 6F&G were repeated with CIITA OE Renca cells. All cell quantifications were by flow cytometry.

As previously demonstrated, WT Renca tumor growth was suppressed by CD4 depletion (Fig. 6F), which decreased Tregs and increased tumor-specific CD8 + lymphocyte activation (Fig. 6G). This is consistent with CD4 depletion generating an antitumor immune response.1 However, for OE Renca, CD4 depletion increased tumor growth (Fig. 6H) and suppressed the antitumor immune response. This happened despite decrease in Tregs, and tumor-specific CD8 + lymphocyte activation was decreased (Fig. 6I). This is consistent with CD4 depletion decreasing antitumor immunity. Therefore, the response to CD4 depletion was similar to what was seen for WT Hepa1-6, which have high baseline MHC II expression.

In TCGA, MHC status identifies tumors where CD4 levels predicts survival

Based on our mechanistic studies, we predicted that patient MHC I and MHC II status will identify subgroups where CD4 levels will predict survival. CD8 levels from bulk tumor RNA sequencing serve as a surrogate for tumor-infiltration of CD8 + lymphocytes and immune activation, and high CD8 expression has been shown to predict better survival.19 However, we hypothesized that tumors with high MHC II expression will be more dependent on CD4 + lymphocytes for suppression of tumor growth and that for these tumors, CD4 expression will predict better survival.

An analysis of the TCGA pan-cancer dataset supported our hypothesis. We illustrate our observation by focusing on two subgroups where our predictions may be most pronounced (Fig. 7). In patients with high MHC II and low MHC I, high CD4 correlated with better progression free survival (p = 0.0217) and trended towards better overall survival (p = 0.0770). Conversely, in patient with low MHC II and high MHC I, high CD4 predicted survival in the opposite direction; high CD4 correlated with worse overall survival (p = 0.0101) and trended towards worse progression free survival (p = 0.1497). Forest plots in Fig. 7 and Supplemental Fig. 7 show how CD4 and CD8 predict overall survival in all the subgroups. The only two subgroups where high CD4 correlated with better survival had high or intermediate MHC II levels.

Fig. 7.

Fig. 7

Open in a new tab

CD4 + status as a predictor of overall survival. (A) Scatterplot based on z-score for MHC I and MHC II expression in all patients in the TCGA pan-cancer dataset. The red lines define the 33.3 and 66.6 percentiles for MHC I and MHC II expression levels. For the four groups defined by the highest and lowest expressions, forest plots show the mean hazard ratio (HR) and 95% confidence interval for the association between CD4 or CD8 T-cell scores and overall survival. Group B (top left corner) has high MHC II and low MHC I expressions. Group C (bottom right corner) has low MHC II and high MHC I expressions. (B-C) Kaplan–Meier curves for overall (OS) and progression-free survival (PFS). Cohorts were divided into two groups using the median CD4 expression as the cutoff. B) In group B, the difference in OS approached significance (p = 0.0770) while the difference in PFS was significant (p = 0.0217), favoring high CD4 expression. (C) In group C, the difference in OS was significant (p = 0.0101), while the difference in PFS approached significance (p = 0.1497), favoring low CD4 expression.

Discussion

This study highlights the importance of patient selection for immunotherapies in general and for CD4 depletion more specifically, underscoring the complexity of the interplay between tumor and immunity. Our findings from various syngeneic mouse tumor models reveal that the effect of CD4 depletion is contingent on the tumor’s MHC expression profile. The suppression of B16 and Renca tumors upon CD4 depletion contrasts with the accelerated growth observed in Hepa1-6 tumors. The causative relation between MHC II and CD4 cells in generating antitumor immunity was established by manipulating MHC II expression in tumors with varying baseline MHC II expressions. We report that in tumors with high MHC II, CD4 + lymphocytes suppress tumor growth while in tumors with low MHC II, CD4 + lymphocytes have the opposite effect on tumor growth.

These observations then led to predictions about how tumor CD4 levels will differentially predict survival in cancer patients with high vs. low tumor MHC II expressions. An analysis of the TCGA pan-cancer dataset bore out our prediction, suggesting that MHC II may be a valuable biomarker for future deployment of CD4 depletion as a novel checkpoint inhibitor. In tumors with high MHC I expression, higher CD8 expression predicted better survival as expected. However, in tumors with low MHC I expression and high MHC II expression, higher CD4 predicted improved survival. In tumors with the opposite profile, that is high MHC I and low MHC II, CD4 predicted survival in the oppositive direction, with higher CD4 predicting worse survival.

Our previous work in B16 and Renca models showed that CD4 depletion removes Tregs, which in turn results in a robust antitumor immune response.1 Zhang et al. showed that CD4 depletion results in protective tumor-specific CD8 memory T-cells that persisted for as long as 5 months following surgical excision of B16 tumors. 4 Using a mouse fibrosarcoma model, Yu et al. showed that CD4 depletion enhances antitumor immunity.5 These studies also showed that the timing of CD4 depletion is important. Optimal results occur with CD4 depletion performed after immune priming, when the majority of CD4 + cells are expected to be immunosuppressive Tregs rather than CD4 + effector cells. 1,4 Consequently, CD4 depletion prior to immune priming can inhibit antitumor immunity. However, from a practical standpoint, this is a clinical scenario that isn’t encountered in patients; by the time a patient’s cancer is diagnosed, the tumor has already primed the immune system.

Based on these preclinical observations, Shitara et al. conducted a phase I clinical trial using a defucosylated humanized anti-CD4 antibody in 11 patients with advanced solid tumors.3 CD8 + T cells increased and one patient with metastatic colon cancer achieved a durable partial response and two additional patients with gastric or esophageal cancer achieved stable disease lasting more than 3 months. In a follow up study, the authors concluded that CD4 depletion is accompanied by expansion of tumor-reactive T-cell clones that mediate antitumor immunity.2 The treatment was well tolerated with only grade I and II toxicities3, which is consistent with prior studies of CD4 depletion for other indications. In several studies, chimeric αCD4 antibodies effectively depleted CD4 levels in cases of refractory cutaneous T-cell lymphoma, without causing serious infections or dose-related toxicities.20–22 A similar CD4 depleting chimeric monoclonal antibody was tested in two randomized, double-blind, placebo-controlled studies for rheumatoid arthritis.23,24 While it didn’t show therapeutic benefits for rheumatoid arthritis, there were no significant adverse events, and patients didn’t experience opportunistic infections.

Tregs are the master regulators of immunity and FoxP3 is the most reliable Treg marker. FoxP3 is specific for Tregs and is required for its function.25 Unfortunately, there is no clinical strategy for targeting FoxP3 expressing cells in patients. Therefore, prior strategies have targeted CD25, which is expressed by the majority of Tregs. However, an important limitation of this approach is that some Tregs are CD25 negative. Furthermore, activated CD8 + T cells express CD25 and can be depleted by CD25-targeting strategies. In murine models, depleting CD25 expressing cells with αCD25 antibodies was effective in preventing tumor growth, but was not effective in treating established tumors 26–28 and has been shown to restrict adoptive immunotherapy 29,30. It is also interesting that daclizumab, an αCD25 antibody, was approved by the FDA for immune suppression and prevention of acute rejection of kidney transplant, which is consistent with its limited role as an immune stimulating anticancer therapy.

CD8 + cytotoxic T cells can kill tumor cells directly and have been the focus of much attention in cancer immunology. Tumor-infiltrating CD8 + lymphocytes are required to generate an effect antitumor immunity, and tumor CD8 + lymphocytes have been proposed as a predictive biomarker and as an early-response marker for cancer immunotherapy.19 However, this and other studies point to the importance of CD4 + T cells.31–37 For example, Ott et al. reported that when administering personalized neoantigen vaccines for melanoma, the MHC II responses were higher than MHC I responses.37 In our TCGA analysis, CD4 and CD8 levels had varying prognostic significant in groups defined by MHC I and MHC II levels. In most groups, high CD4 correlated with worse survival since CD4 + cells are likely Tregs. However, when MHC II was high and MHC I was low, high tumor CD4 levels predicted better survival since CD4 + cells are likely effectors cells contributing to antitumor immunity.

Our findings have important implications for development of CD4 depletion therapy. Future clinical trials of CD4 depletion strategies should consider MHC status as a predictive biomarker, and if validated as a predictor of treatment response, MHC status should be used for patient selection. The ability to enrich for patients most likely to respond to a therapy makes a treatment highly attractive. Also, when planning for phase II testing, baseline MHC I and MHC II status can help prioritize tumor types for clinical testing. Finally, additional preclinical studies should assess the relationship between lymphocytes, MHC status and myeloid cells.

In conclusion, these observations highlight the role of CD4 + lymphocytes in controlling or promoting cancer growth, and identify MHC status as a determinant of these divergent roles. In the TCGA pan-cancer dataset, we found corroborative evidence suggesting that MHC status can determine whether CD4 or CD8 may be the best tissue-based early-response marker for immunotherapy. Future clinical trials of CD4 depleting cancer immunotherapy should evaluate MHC II as a biomarker for patient selection.

Supplementary Information

Supplementary Information. (14MB, pptx)

Author contributions

H.K. wrote the main manuscript text. Y.W. and S.S. performed immune experiments and data analysis. M.K. and S.Y. performed data analysis with the TCGA dataset. L.H. performed animal experiments and data analysis. All authors contributed to study conception and design, and reviewed the manuscript.

The first draft of the manuscript was written by Hyung Kim and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data availability

The raw experimental data generated during the current study are available from the corresponding author on reasonable request. Human data used in the current study are publicly available.

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-88972-8.

References

  • 1.Wang, Y., Sparwasser, T., Figlin, R. & Kim, H. L. Foxp3+ T cells inhibit antitumor immune memory modulated by mTOR inhibition. Cancer Res.74, 2217–2228. 10.1158/0008-5472.CAN-13-2928 (2014). [DOI] [PubMed] [Google Scholar]
  • 2.Aoki, H. et al. Transient depletion of CD4(+) cells induces remodeling of the TCR repertoire in gastrointestinal cancer. Cancer Immunol. Res.9, 624–636. 10.1158/2326-6066.CIR-20-0989 (2021). [DOI] [PubMed] [Google Scholar]
  • 3.Shitara, K. et al. First-in-human phase 1 study of IT1208, a defucosylated humanized anti-CD4 depleting antibody, in patients with advanced solid tumors. J. Immunother. Cancer7, 195. 10.1186/s40425-019-0677-y (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Zhang, P., Cote, A. L., de Vries, V. C., Usherwood, E. J. & Turk, M. J. Induction of postsurgical tumor immunity and T-cell memory by a poorly immunogenic tumor. Cancer Res.67, 6468–6476. 10.1158/0008-5472.CAN-07-1264 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Yu, P. et al. Intratumor depletion of CD4+ cells unmasks tumor immunogenicity leading to the rejection of late-stage tumors. J. Exp. Med.201, 779–791. 10.1084/jem.20041684 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Ostrand-Rosenberg, S. et al. Cell-based vaccines for the stimulation of immunity to metastatic cancers. Immunol. Rev.170, 101–114. 10.1111/j.1600-065x.1999.tb01332.x (1999). [DOI] [PubMed] [Google Scholar]
  • 7.Ostrand-Rosenberg, S., Thakur, A. & Clements, V. Rejection of mouse sarcoma cells after transfection of MHC class II genes. J. Immunol.144, 4068–4071 (1990). [PubMed] [Google Scholar]
  • 8.Percie, S. N. et al. Reporting animal research: Explanation and elaboration for the ARRIVE guidelines 20. PLoS Biol18, e000411. 10.1371/journal.pbio.3000411 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Goldman, M. J. et al. Visualizing and interpreting cancer genomics data via the Xena platform. Nat. Biotechnol.38, 675–678. 10.1038/s41587-020-0546-8 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Wang, S. & Liu, X. The UCSCXenaTools R package: A toolkit for accessing genomics data from UCSC Xena platform, from cancer multi-omics to single-cell RNA-seq. J. Open Source Softw.4, 1627. 10.21105/joss.01627 (2019). [Google Scholar]
  • 11.Levine, D. M. et al. Pathway and gene-set activation measurement from mRNA expression data: the tissue distribution of human pathways. Genome Biol.7, R93. 10.1186/gb-2006-7-10-r93 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Racle, J., de Jonge, K., Baumgaertner, P., Speiser, D. E. & Gfeller, D. Simultaneous enumeration of cancer and immune cell types from bulk tumor gene expression data. eLife6, e26476. 10.7554/eLife.26476 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Sturm, G. et al. Comprehensive evaluation of transcriptome-based cell-type quantification methods for immuno-oncology. Bioinformatics35, i436–i445. 10.1093/bioinformatics/btz363 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Hiotis, S. P., Yee, H. T., Luan, W. & Burakoff, S. J. Site does matter: tumorigenesis of the murine hepatocellular carcinoma cell line Hepa 1–6 is dependent upon intrahepatic rather than subdermal implantation. J. Surg. Res.114, 274 (2003). [Google Scholar]
  • 15.He, Y. et al. MHC class II expression in lung cancer. Lung Cancer112, 75–80. 10.1016/j.lungcan.2017.07.030 (2017). [DOI] [PubMed] [Google Scholar]
  • 16.Ruan, X. et al. Reduced MHC class II expression in medullary thyroid cancer identifies patients with poor prognosis. Endocr. Relat. Cancer29, 87–98. 10.1530/ERC-21-0153 (2022). [DOI] [PubMed] [Google Scholar]
  • 17.Forero, A. et al. Expression of the MHC class II Pathway in triple-negative breast cancer tumor cells is associated with a good prognosis and infiltrating lymphocytes. Cancer Immunol. Res.4, 390–399. 10.1158/2326-6066.CIR-15-0243 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Wu, B., Wang, Q., Li, B. & Jiang, M. LAMTOR1 degrades MHC-II via the endocytic in hepatocellular carcinoma. Carcinogenesis43, 1059–1070. 10.1093/carcin/bgac075 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Li, F. et al. The association between CD8+ tumor-infiltrating lymphocytes and the clinical outcome of cancer immunotherapy: A systematic review and meta-analysis. EClinicalMedicine41, 101134. 10.1016/j.eclinm.2021.101134 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Kim, Y. H. et al. Clinical efficacy of zanolimumab (HuMax-CD4): two phase 2 studies in refractory cutaneous T-cell lymphoma. Blood109, 4655–4662. 10.1182/blood-2006-12-062877 (2007). [DOI] [PubMed] [Google Scholar]
  • 21.d’Amore, F. et al. Phase II trial of zanolimumab (HuMax-CD4) in relapsed or refractory non-cutaneous peripheral T cell lymphoma. Br. J. Haematol.150, 565–573. 10.1111/j.1365-2141.2010.08298.x (2010). [DOI] [PubMed] [Google Scholar]
  • 22.Knox, S. et al. Treatment of cutaneous T-cell lymphoma with chimeric anti-CD4 monoclonal antibody. Blood87, 893–899 (1996). [PubMed] [Google Scholar]
  • 23.Moreland, L. W. et al. Double-blind, placebo-controlled multicenter trial using chimeric monoclonal anti-CD4 antibody, cM-T412, in rheumatoid arthritis patients receiving concomitant methotrexate. Arthritis and rheumatism38, 1581–1588 (1995). [DOI] [PubMed] [Google Scholar]
  • 24.van der Lubbe, P. A., Dijkmans, B. A., Markusse, H. M., Nassander, U. & Breedveld, F. C. A randomized, double-blind, placebo-controlled study of CD4 monoclonal antibody therapy in early rheumatoid arthritis. Arthritis and rheumatism38, 1097–1106 (1995). [DOI] [PubMed] [Google Scholar]
  • 25.Gavin, M. A. et al. Foxp3-dependent programme of regulatory T-cell differentiation. Nature445, 771–775. 10.1038/nature05543 (2007). [DOI] [PubMed] [Google Scholar]
  • 26.Quezada, S. A. et al. Limited tumor infiltration by activated T effector cells restricts the therapeutic activity of regulatory T cell depletion against established melanoma. J. Exp. Med.205, 2125–2138. 10.1084/jem.20080099 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Elpek, K. G., Lacelle, C., Singh, N. P., Yolcu, E. S. & Shirwan, H. CD4+CD25+ T regulatory cells dominate multiple immune evasion mechanisms in early but not late phases of tumor development in a B cell lymphoma model. J. Immunol.178, 6840–6848 (2007). [DOI] [PubMed] [Google Scholar]
  • 28.Onizuka, S. et al. Tumor rejection by in vivo administration of anti-CD25 (interleukin-2 receptor alpha) monoclonal antibody. Cancer Res.59, 3128–3133 (1999). [PubMed] [Google Scholar]
  • 29.Curtin, J. F. et al. Treg depletion inhibits efficacy of cancer immunotherapy: Implications for clinical trials. PLoS One3, e1983. 10.1371/journal.pone.0001983 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Cote, A. L., Byrne, K. T., Steinberg, S. M., Zhang, P. & Turk, M. J. Protective CD8 memory T cell responses to mouse melanoma are generated in the absence of CD4 T cell help. PLoS One6, e26491. 10.1371/journal.pone.0026491 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Wang, R. F. The role of MHC class II-restricted tumor antigens and CD4+ T cells in antitumor immunity. Trends Immunol.22, 269–276. 10.1016/s1471-4906(01)01896-8 (2001). [DOI] [PubMed] [Google Scholar]
  • 32.Kim, J. Y. et al. MHC II immunogenicity shapes the neoepitope landscape in human tumors. Nat. Genet.55, 221–231. 10.1038/s41588-022-01273-y (2023). [DOI] [PubMed] [Google Scholar]
  • 33.Huang, L. et al. Small-molecule MHC-II inducers promote immune detection and anti-cancer immunity via editing cancer metabolism. Cell Chem Biol30, 1076–1089. 10.1016/j.chembiol.2023.05.003 (2023). [DOI] [PubMed] [Google Scholar]
  • 34.Mortara, L. et al. CIITA-induced MHC class II expression in mammary adenocarcinoma leads to a Th1 polarization of the tumor microenvironment, tumor rejection, and specific antitumor memory. Clin Cancer Res12, 3435–3443. 10.1158/1078-0432.CCR-06-0165 (2006). [DOI] [PubMed] [Google Scholar]
  • 35.Meazza, R., Comes, A., Orengo, A. M., Ferrini, S. & Accolla, R. S. Tumor rejection by gene transfer of the MHC class II transactivator in murine mammary adenocarcinoma cells. Eur. J. Immunol.33, 1183–1192. 10.1002/eji.200323712 (2003). [DOI] [PubMed] [Google Scholar]
  • 36.Panelli, M. C. et al. Interferon gamma (IFNgamma) gene transfer of an EMT6 tumor that is poorly responsive to IFNgamma stimulation: increase in tumor immunogenicity is accompanied by induction of a mouse class II transactivator and class II MHC. Cancer Immunol. Immunother.42, 99–107. 10.1007/s002620050258 (1996). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Ott, P. A. et al. An immunogenic personal neoantigen vaccine for patients with melanoma. Nature547, 217–221. 10.1038/nature22991 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information. (14MB, pptx)

Data Availability Statement

The raw experimental data generated during the current study are available from the corresponding author on reasonable request. Human data used in the current study are publicly available.

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES