Dysregulation of Gene Expression of Key Signaling Mediators in PBMCs from People with Type 2 Diabetes Mellitus

Diabetes is currently the fifth leading cause of death by disease in the USA. The underlying mechanisms for type 2 Diabetes Mellitus (DM2) and the enhanced susceptibility of such patients to inflammatory disorders and infections remain to be fully defined. We have recently shown that peripheral blood mononuclear cells (PBMCs) from non-diabetic people upregulate expression of inflammatory genes in response to proteasome modulators, such as bacterial lipopolysaccharide (LPS) and soybean lectin (LEC); in contrast, resveratrol (RES) downregulates this response. We hypothesized that LPS and LEC will also elicit a similar upregulation of gene expression of key signaling mediators in (PBMCs) from people with type 2 diabetes (PwD2, with chronic inflammation) ex vivo. Unexpectedly, using next generation sequencing (NGS), we show for the first time, that PBMCs from PwD2 failed to elicit a robust LPS- and LEC-induced gene expression of proteasome subunit LMP7 (PSMB8) and mediators of T cell signaling that were observed in non-diabetic controls. These repressed genes included: PSMB8, PSMB9, interferon-γ, interferon-λ, signal-transducer-and-activator-of-transcription-1 (STAT1), human leukocyte antigen (HLA DQB1, HLA DQA1) molecules, interleukin 12A, tumor necrosis factor-α, transporter associated with antigen processing 1 (TAP1), and several others, which showed a markedly weak upregulation with toxins in PBMCs from PwD2, as compared to those from non-diabetics. Resveratrol (proteasome inhibitor) further downregulated the gene expression of these inflammatory mediators in PBMCs from PwD2. These results might explain why PwD2 may be susceptible to infectious disease. LPS and toxins may be leading to inflammation, insulin resistance, and thus, metabolic changes in the host cells.


Introduction
Diabetes mellitus (DM) is currently the fifth leading cause of death by disease in the U.S. As of 2015,~415 million individuals worldwide (8.3% of adults) have been diagnosed with diabetes. Many diabetes investigators have focused on reducing insulin resistance and reducing elevated circulating glucose levels to impact the complications, morbidity, and mortality of people with type 2 diabetes mellitus (PwD2), which generally develop later in life and are associated with diet and lifestyle. The role of PwD2 in increased susceptibility and worse outcomes of infectious diseases has been well documented [1][2][3][4][5][6][7][8][9]. It has also been implicated in the higher risks and mortality of viral diseases, such as COVID-19 [1,2], AIDS, SARS, Herpes Simplex, and enterovirus. The first line of treatment for PwD2 is metformin, as suggested by the American Diabetes Association ADA [10,11]. However, the Int. J. Mol. Sci. 2023, 24, 2732 3 of 25 because it can induce gene expression of macrophage and T-cell cytokines IFN-γ, IL-4, IL-2, and LMP7 (PSMB8) in non-diabetic controls just like LPS. The mechanisms by which plant products, such as RES (anti-inflammatory compound) and LEC (inflammatory compound), affect the gene expression of mediators in signaling pathways in human PBMCs of PwD2 are presently not understood [41][42][43][44][45][46][47][48][49][50][51][52][53][54].
We have recently reported that RES and LEC differentially affect LPS-induced pathways in non-diabetic controls via repression or induction of PSMB8, respectively [54]. Our present study was based on the hypothesis that PBMCs from non-diabetics and PwD2 would be differentially responsive to inducing the gene expression of mediators involved in signaling pathways with LPS (bacterial) and plant soybean lectins (lectins are also present in plants, bacteria, and viruses) ex vivo, being metabolically different. This would lead to an alternative modulation in gene expression of insulin receptors, glucose transporters, proteasome subunits, and mediators involved in the signaling pathways (and perhaps leading to insulin-resistance) and causing disease. Therefore, to test this hypothesis, we determined the extent to which proteasome modulators, RES and LEC, play a role in regulating the gene expression of mediators of signaling pathways in PBMCs from non-diabetics and PwD2, in the presence and absence of LPS.

Next Generation Sequencing (NGS) Profiling Experiment
PBMCs from non-diabetic controls [54] and PwD2 (age, weight, and sex-matched) were treated with or without LPS, RES, LEC10, and LEC50. PBMCs were treated with agonists to investigate if cells from PwD2 would be able to respond to LEC, LPS and RES similarly to the non-diabetic control [54]. RNA was extracted and subjected to NGS, and log2 fold ratios were uploaded onto the Ingenuity Pathway Analysis Program (Qiagen). The two objectives of this study were as follows: 1. To define gene expression differences in PBMCs between non-diabetic and PwD2 (D). The control values mentioned in our previous manuscript were used [54]; 2. To define differences in gene expression in PBMCs from PwD2 treated with proteasome modulators, LPS (DLPS), LEC (DLEC), LPS + RES (DLPS + RES), and RES (DRES) alone using NGS. Table 1 shows the number of differentially expressed genes (DEG) in RNA samples. All the genes induced/repressed are summarized in Table 1, all other tables are presented in the Supplementary Materials (SM). LPS induced 555 genes in PBMCs from non-diabetic controls, but only 377 in PwD2. Total genes modulated by RES (80 µM), LPS (10 ng/mL), and LPS + RES (10 ng/mL and 80 µM), LEC10 (10 µg/mL). LEC10 + LPS, LEC50 (50 µg/mL) and LEC50 + LPS in PBMCs of non-diabetic controls and PwD2 are listed in Table 1.

Controls and PwD2 with LPS, RES, and LEC
It is well-established that PwD2 have increased levels of cytokines, including TNF-α, IL-6, and IL-18 [13][14][15][16][17]. We investigated the cell functions that were upregulated by LPS and downregulated by RES in PBMCs from non-diabetics RES (R), LPS (L), and LPS-RES (LR), PwD2 RES (DR), PwD2 LPS (DL), and PwD2-LPS-RES (DLR). Respective controls for non-diabetic controls and PwD2 were used. Importantly, functional analysis of genes as related to the activation of phagocytes, blood cells, leukocytes, induction of cells, chemotaxis, homing of cells, inflammatory response, migration, cell movement, and all others ( Figure 1) were robustly downregulated in RNA obtained from PBMCs of PwD2, as compared to the non-diabetic controls after treatment with LPS. RES further downregulated these genes, while LEC and LPS upregulated gene expression of genes, as described below. Human PBMCs from non-diabetic and PwD2 were treated with three different treatments and the vehicle control for 3 h. The RNA samples from controls and PwD2 were extracted from the cells and analyzed using RNAseq. These data were first extracted using the DEG analysis, followed by ingenuity pathways analysis. Cell functions for RES (R), LPS (L), LPS + RES (L + R), DRES (DR), DLPS (DL), and DLPS + RES (DL + R). Comparisons were made using their respective controls. Dark orange color denotes maximal effect, while dark blue color denotes minimal effect.
Normally, PwD2 have been reported to have excessive inflammation in their cells followed by insulin-resistance, which leads to heart disease, cancer, and wound healing. Importantly, failure of LPS or LEC to elicit a robust upregulation of expression of genes that belonged to mediators of signaling pathways, such as pattern recognition, cardiovascular signaling, AMPK, Th17 activation, Th1 (IFN-γ), Th2 (IL-4), STAT3, amyotrophic lateral sclerosis and death receptor signaling, was observed in PBMCs of PwD2, as compared to non-diabetic controls (some shown in Figure 2A-F). LPS also showed down-regulation of gene expression of mediators responsible for acute phase signaling, antioxidant action of vitamin C, and endocannabinoid cancer inhibition pathways. In contrast, gene expression of mediators involved in opioid signaling was upregulated by LPS in PBMCs of PwD2, as compared to controls.
(AHRR), and G-protein coupled receptor 68 (GPR68). The expression of both inflammatory and anti-inflammatory mediator genes was downregulated when RES alone was used (a proteasome CT-like inhibitor of subunit LMP7). However, some genes were also upregulated by RES, as compared to PBMCs from PwD2, as shown in Table S3B.
LPS upregulated gene expression of mediators for most disease functions, whereas RES inhibited this, as shown in Figure 1. We observed LPS-induced changes in gene expression of human leukocyte antigens (antigen presentation, HLA molecules).

Further downregulation of gene expression in mediators of inflammatory signaling pathways in PBMCs from PwD2 was observed by RES treatment (alone).
To analyze RES-modulated DEG analysis, we have included inflammatory marker genes that were downregulated as shown in Table S3A. Several of the significant inflammation-linked genes were downregulated with RES alone, these include chemokine genes (CCL2 or MCP1), thrombospondin 1 (THB1), and thrombomodulin (THBD); cytochrome P450 family (CYP1B1), early growth response 2 (EGR2), chemokine (CCL7), cannabinoid receptor 2, (CNR2), formyl peptide receptor 2 (FPR2), IL-10, aryl-hydrocarbon receptor repressor (AHRR), and G-protein coupled receptor 68 (GPR68). The expression of both inflammatory and anti-inflammatory mediator genes was downregulated when RES alone was used (a proteasome CT-like inhibitor of subunit LMP7). However, some genes were also upregulated by RES, as compared to PBMCs from PwD2, as shown in Table S3B.
LPS upregulated gene expression of mediators for most disease functions, whereas RES inhibited this, as shown in Figure 1. We observed LPS-induced changes in gene expression of human leukocyte antigens (antigen presentation, HLA molecules). Although gene expression of several mediators in canonical pathways was inhibited by RES, but we have focused on IFN-γ, Th1, Th2, and pattern-recognition pathways in this manuscript (Figure 2A-F) because those were affected the most. In contrast, gene expression for key mediators in pathways for the synthesis of prostaglandins and fatty acid synthesis was upregulated by RES in human PBMC. To investigate the signaling pathways affected by RES (DR), LPS (DL), and LPS + RES (DLR), we found that LPS upregulated genes of key mediators of several inflammatory pathways that were robustly downregulated by treatment with RES in PBMCs of PwD2. These signaling pathways also included Th17 activation, HMG-B1, P38 MAPK, Th1 (IFN-γ), and Th2 (IL-4), triggering the receptor expressed on myeloid cells 1 (TREM1), IL-6, and neuroinflammation, as shown in Figure 3. Although gene expression of several mediators in canonical pathways was inhibited by RES, but we have focused on IFN-γ, Th1, Th2, and pattern-recognition pathways in this manuscript (Figure 2A-F) because those were affected the most. In contrast, gene expression for key mediators in pathways for the synthesis of prostaglandins and fatty acid synthesis was upregulated by RES in human PBMC. To investigate the signaling pathways affected by RES (DR), LPS (DL), and LPS + RES (DLR), we found that LPS upregulated genes of key mediators of several inflammatory pathways that were robustly downregulated by treatment with RES in PBMCs of PwD2. These signaling pathways also included Th17 activation, HMG-B1, P38 MAPK, Th1 (IFN-γ), and Th2 (IL-4), triggering the receptor expressed on myeloid cells 1 (TREM1), IL-6, and neuroinflammation, as shown in Figure  3.  Figure 2. RNAseq data were first extracted using the DEG analysis and analyzed by ingenuity pathways analysis. Z scores were plotted against the signaling pathways, where dark orange denotes maximum activation and dark blue maximum repression. Figure 4A). The modulation of most of these above-mentioned genes may be dependent on the UPS because its proteases degrade many of the short-lived proteins. To investigate major differences in proteasome subunits, PBMCs from LPS-treated controls and PwD2 were compared in Figure   We have previously shown that the proteasome is a central regulator of inflammation and cellular function by LPS. It contains the six protease subunits X, Y and Z, and the inducible subunits LMP7, LMP2, and LMP10. The PBMCs contain proteasome's protease genes, predominantly subunits LMP7, LMP2, Y, and Z, while X and LMP10 were not significantly expressed in PBMCs (all six subunits are usually expressed in mouse C57Bl/6 macrophages and predominantly X, Y, and Z in RAW 264.7 cells). LPS and LEC50 upreg- We have previously shown that the proteasome is a central regulator of inflammation and cellular function by LPS. It contains the six protease subunits X, Y and Z, and the inducible subunits LMP7, LMP2, and LMP10. The PBMCs contain proteasome's protease genes, predominantly subunits LMP7, LMP2, Y, and Z, while X and LMP10 were not significantly expressed in PBMCs (all six subunits are usually expressed in mouse C57Bl/6 macrophages and predominantly X, Y, and Z in RAW 264.7 cells). LPS and LEC50 upregulated the gene expression of PSMB8 (LMP7) and PSMB9 (LMP2) in non-diabetic PBMC, but to a lesser extent in those from PwD2 ( Figure 4A,B), but PSMB6 (Y) and PSMB7 (Z) were downregulated in both controls and PwD2 ( Figure 4C,D). RES downregulated the LPS-induced gene expression of PSMB8 and PSMB9 in PBMCs of both PwD2 and controls.
LEC50 alone also showed less robust expression of genes of key mediators involved in several signaling pathways in PBMC from PwD2, as compared to non-diabetic controls. Several important genes were also downregulated by LEC50, as listed in Table S5B. These genes included serine dehydratase (SDS, the deamination of L-serine to yield pyruvate), heme-oxygenase 1 (HMOX1, degrades heme to biliverdin and bilirubin), CD163 (glycoprotein, is a sensor for Gram-negative and Gram-positive bacteria, scavenger receptor for hemoglobin-haptoglobin complex), fibrinogen (FGL2 induces blood clots, but the deficiency causes bleeding and thrombosis), and chemokine receptor 1 (CCR1). Other downregulated genes include pyruvate dehydrogenase kinase isozyme 4 (PDK4), CD14 (LPS binding protein), claudin 5 (CLDN5, integral membrane proteins and components of tight junction strands in endothelial cells), insulin receptor (INSR binds insulin), NRL family CARD domain (NRLC4), vitamin D3 receptor (VDR), N-acetylneuraminate pyruvate lyase (NPL, N-acetyl-D-mannosamine and pyruvate are products of this enzyme), and lysozyme (LYZ, antimicrobial enzyme). This suggests that LEC50 does not elicit robust upregulation of several genes that are important for Th1 immune response to infections in PBMC from PwD2, especially IFN-γ. Table S6 shows the modulation of transcription factor genes by various treatments of PBMCs from non-diabetic controls vs. PwD2.
LEC50 downregulates gene expression of mediators involved in four major pathways, and these include: PPAR signaling, antioxidant action of vitamin C, LXR/RXR activation (Figure 2), and Gαι signaling. LPS induces IFN-γ, IL-4 in PBMCs in nondiabetic controls (CLPS), but to a lesser extent in PwD2 (DLPS). Importantly, soybean LEC50 upregulated the expression of many of the genes that toxic LPS does in non-diabetic controls [54] (Figure 6, CLEC10, CLEC50, blue bars), but not to that extent in PwD2 (LEC10, LEC50, red bars). However, LEC10 and LEC50 can reverse this effect by inducing the gene expression of some of these cytokines in PBMCs from PwD2. Several other cytokines were modulated differentially, as shown in Table 2, Figure 6.    The NGS data for the gene expression for two cytokines, TNF-α and IFN-γ, was validated by qPCR using PBMC from non-diabetic vs. PwD2 in four additional experiments and the results were like those obtained in this study (Figure 7). The data and the trend with proteasome modulators were reproducible. A minimal upregulation of gene expression of TNF-α and IFN-γ was observed. LEC upregulated the gene expression of TNFα and IFN-γ dose-dependently in the PBMCs of non-diabetics, but not in those from PwD2 (Figures 7 and 8). Moreover, LEC10 upregulated the gene expression of IL-2 more robustly in PBMCs from PwD2, as compared to IFN-γ (Figure 7). The gene expression of iNOS and TNF-α was downregulated in PBMCs from PwD2, as compared to non-diabetic controls, as shown in Figure 8. The NGS data for the gene expression for two cytokines, TNF-α and IFN-γ, was validated by qPCR using PBMC from non-diabetic vs. PwD2 in four additional experiments and the results were like those obtained in this study (Figure 7). The data and the trend with proteasome modulators were reproducible. A minimal upregulation of gene expression of TNF-α and IFN-γ was observed. LEC upregulated the gene expression of TNF-α and IFN-γ dose-dependently in the PBMCs of non-diabetics, but not in those from PwD2 (Figures 7 and 8). Moreover, LEC10 upregulated the gene expression of IL-2 more robustly in PBMCs from PwD2, as compared to IFN-γ (Figure 7). The gene expression of iNOS and TNF-α was downregulated in PBMCs from PwD2, as compared to non-diabetic controls, as shown in Figure 8.  Gene expression of LPS-induced STAT1 and IFN-γ was not robustly upregulated in PBMCs from PwD2 as compared to those from non-diabetic controls. PBMCs from non-diabetic controls show a robust upregulation of the gene expression response to LPS. Gene expression levels of IFN-γ, STAT1, STAT2, and many of the downstream proteins were upregulated with LPS in non-diabetic controls ( Figure 9A) [54]. In contrast, PBMCs from PwD2 failed to elicit a robust gene expression response to LPS, and a significant upregulation in the gene expression of IFN-γ signaling pathways via STAT1 ( Figure 9B).

Figure 8.
PBMCs were treated with either vehicle, RES, LPS, LPS + RES, or LEC10. RNA was extracted from cells and gene expression of IFN-γ and TNF-α was analyzed using qRT-PCR. Mean+ SEM. In these graphs, C, non-diabetic control was used to calculate 2 −ΔΔCT for non-diabetics, and D control was used for diabetics. LPS did not induce gene expression of IFN-γ as robustly in PBMCs from PwD2, compared to non-diabetic controls. Values in a column sharing a common asterisk with proteasome modulators were significantly different at *, **, *** p < 0.024, 0.007, 0.0001, using 1-way Anova. 2 -ΔΔCT = Relative quantification RQ. (Experiments 1-4 are described in the Methods section).

Gene expression of LPS-induced STAT1 and IFN-γ was not robustly upregulated in PBMCs from
PwD2 as compared to those from non-diabetic controls. PBMCs from non-diabetic controls show a robust upregulation of the gene expression response to LPS. Gene expression levels of IFN-γ, STAT1, STAT2, and many of the downstream proteins were upregulated with LPS in non-diabetic controls ( Figure 9A) [54]. In contrast, PBMCs from PwD2 failed to elicit a robust gene expression response to LPS, and a significant upregulation in the gene expression of IFN-γ signaling pathways via STAT1 ( Figure 9B). Figure 8. PBMCs were treated with either vehicle, RES, LPS, LPS + RES, or LEC10. RNA was extracted from cells and gene expression of IFN-γ and TNF-α was analyzed using qRT-PCR. Mean ± SEM. In these graphs, C, non-diabetic control was used to calculate 2 −∆∆CT for non-diabetics, and D control was used for diabetics. LPS did not induce gene expression of IFN-γ as robustly in PBMCs from PwD2, compared to non-diabetic controls. Values in a column sharing a common asterisk with proteasome modulators were significantly different at *, **, *** p < 0.024, 0.007, 0.0001, using 1-way Anova. 2 −∆∆CT = Relative quantification RQ. (Experiments 1-4 are described in the Section 4).

Discussion
Mounting a good immune response (upregulation of IFN-γ) to viral or bacterial toxins is essential for the proper disposal of infectious agents, as observed in PBMCs from nondiabetic controls [54]. Proteasomes play a central role in most of the functions of the cell via the degradation of proteins. Proteasomes utilize their six proteases to degrade unwanted proteins in the cell, signaling proteins, transcription factors, cell-cycle proteins, cytokines, and to control the cellular metabolism at the proper time. Recently, we reported that proteasome activators, LPS and LEC, upregulate LMP7 and several inflammatory mediators, while proteasome inhibitor RES downregulates some of these mediators in PBMCs from non-diabetic controls. Therefore, we wanted to determine if chronically inflamed PBMCs from PwD2 would also react similarly to proteasome modulators, such as LPS, or other toxins, such as soybean LEC and flavonoids (RES), as efficiently as those from non-diabetic controls. Unexpected results were obtained since PBMCs from PwD2 showed a failure to elicit a normal response to LPS and lectins. LPS could not induce the gene expression of IL-12A, HLA complex II, IFN-γ, TNF-α, LMP7, and IFN-λ presumably because PBMCs from PwD2 were already in a more inflamed and insulin-resistant state. This may lead the PBMCs to become tolerant and refractory to LPS as discussed below. We also provide evidence here that dietary components, such as RES (proteasome inhibitor of predominantly subunit LMP7), showed a comparable anti-inflammatory response in the expression of some genes, while LEC showed a normal or reduced induction of gene expression response in PBMCs from PwD2, as compared to non-diabetic controls.
Our data showed that in ex vivo culture, PBMCs from PwD2 were dysregulated immunologically since gene expression of mediators of several pathways and cytokines were affected. These PBMCs were relatively unresponsive to LPS with respect to the gene expression of IFN-γ and key mediators in signaling pathways as compared to non-diabetic controls. IFN-γ belongs to a family of cytokines that function as potent macrophageactivating, microbicidal, and antiviral agents in upregulating host defense mechanisms. Major differences were noted in the DEG analysis of RNA from PBMCs of non-diabetic controls vs. PwD2, only upon further treatment with LPS or LEC.
This manuscript provides strong evidence to support the conclusion that PBMCs from PwD2 show a failure in eliciting a normal response to LPS, as compared to non-diabetic controls. Importantly, this failure in eliciting an upregulation in the gene expression of key mediators in IFN-γ signaling and others ( Figure 6), such as interferon lambda (IFN-λ, involved with dendritic cell function), transporter associated with antigen presentation (TAP1), tumor necrosis factor (TNF-α), signal-transducer and activator of transcription-1 (STAT1), cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), also known as CD152  Table 2.
There are several reasons why PBMCs from PwD2 might have failed to elicit a robust immune response to LPS and LEC with respect to IFN-γ signaling and other pathways. This could be due to some or all of the following factors: (1) Metformin is an anti-inflammatory compound and is widely used for PwD2 [56][57][58]; (2) The low induction of the gene expres-sion of LMP7 (PSMB8), and other subunits of the proteasome; (3) The induction of genes linked to epigenetic pathways; (4) The weak dendritic-T cell interaction; (5) Tolerance due to previous exposure to LPS or lectins, or T-cell senescence due to chronic inflammation; (6) Tolerant PBMCs being in a state where IFN-γ is no longer inducible. Therefore, significant upregulation in the gene expression of the IFN-γ signaling pathways via STAT1 was not observed. Previously, some researchers have reported a slight upregulation of IFN-γ, while others have reported a downregulation in PwD2 [59][60][61][62]. We have shown for the first time, that PBMCs from PwD2 show a dysregulated response to toxins such as LPS and lectins, as described below. Moreover, the repression of several genes was observed in PBMCs from PwD2, which could be due to lower gene expression of two important transcription factors that were downregulated, T-box 3 (TBX3) and neuregulin 1 (NRG) ( Table S6). These transcription factors have previously been shown to be important for cancer and obesity, respectively [63,64]. The net result observed in PBMCs from PwD2 could be due to the weak crosstalk between the dendritic cells (low abundance of CD80 and MHC-II) and T cells.
The upregulated genes in PBMCs of PwD2 (without LPS treatment) included enzymes, such as the mitochondrial isozyme of pyruvate dehydrogenase kinase 4 (PDK4), the increased expression of which is linked to decreased metabolism, hypoxia, conservation of glucose by decreasing acetyl-CoA, (which enters in the citric acid cycle, and produces ATP). ATP is required for proteasomal activation and inflammatory processes because protein degradation by the UPS would shut down without it. It is well established that LPS causes insulin resistance in the mouse, as well as in the human model of DM2 [65][66][67][68]. LPS/LEC causes the downregulation of gene expression of insulin receptor (INSR) and the upregulation of INSRR, INSM1 [69], and TRIP10 in PBMCs from PwD2 and healthy controls, lectins may also be contributing to hormonal changes, such as insulin resistance and thyroid hormone receptor interaction, resulting in metabolic changes in the host.
To place these observations within the framework of what is currently understood regarding LPS-mediated signaling pathways, we propose the following simplistic model for mechanisms involved in human PBMC in PwD2 and non-diabetics in Figure 10. It is well-established that dendritic cells can cross talk with the naïve Th0 CD4 cells using their MHC-II molecules and T cell receptor (TCR) via interaction of CD28 and CD80 (B7). LPS leads to the induction of the gene expression of pro-inflammatory cytokines (cytokine storm) and mediators such as IL-1β, IL-6, IL-12, TNF-α, IFN-β, F3, F8, COX2, and IL-27 in PBMCs. The IFN-γ (from Th1 or natural killer cells, NK cells) induced by agonists and transcription factor STAT1 activate LMP7 subunits of the proteasomes of Th1 cells and macrophages [54]. While Th2 cytokines IL-2, IL-4, IL-5, STAT6, and GATA3 activate B cells (CD19) to induce memory cells and plasma cell antibodies (IgM, IgG, IgA and IgE) in non-diabetic cells. We report for the first time that LPS and LEC50 treatments lead to a robust induction of gene expression of pro-inflammatory cytokines and innate immunity mediators such as IL-1β, IL-6, IL-12B, IFN-β, F3, F8, COX2, and IL-27 in PBMCs from nondiabetic controls and PwD2 (using NGS). In contrast, the gene expression of IL-12A, STAT1, PSMB8, and MHC-II HLA DQB, as well as most IFN-γ-induced proteins and several other key cytokine genes involved in the adaptive immune response observed in non-diabetics, were not upregulated to the same extent in PBMCs of PwD2. PBMCs from PwD2 upon activation with LPS showed minimal gene expression of T cell cytokines with respect to Th1 (IFN-γ), Th2 (IL-4, IL-5), Th17 (IL-17) and Treg cell (IL-10, IL-20) cytokines, except for Th1 (IL-2). IL-2 is known to activate Treg cells, which calms down the inflammatory response [55].
Th1 (IFN-γ), Th2 (IL-4, IL-5), Th17 (IL-17) and Treg cell (IL-10, IL-20) cytokines, except for Th1 (IL-2). IL-2 is known to activate Treg cells, which calms down the inflammatory response [55]. The natural products RES reduced the gene expression of these IFN-γ-signaling cytokine gene response in T cells even further; in contrast, LEC50 showed activation of the same inflammatory gene expression patterns as LPS in PBMCs of controls and PwD2. LEC could not upregulate the gene expression of IFN-γ, PSMB8, CD80, TCR, and MHC II involved in adaptive immune response and T-cell killing to the same extent in PBMCs of PwD2, as compared to non-diabetic controls. These data could explain why PwD2 are more vulnerable to bacterial and viral diseases as compared to non-diabetic individuals. These data also provide support for the hypothesis that bacterial and plant toxins (LPS/bacteria/viruses and lectins from plants) would react differentially to elicit the gene upregulation of important mediators in PBMCs from non-diabetic controls and PwD2.
This study also revealed for the first time that bacterial (LPS) and plant toxins may cause robust inflammation in healthy controls and PwD2, leading to the downregulation of the gene expression of the insulin receptor (INSR). In addition, the gene expression of The natural products RES reduced the gene expression of these IFN-γ-signaling cytokine gene response in T cells even further; in contrast, LEC50 showed activation of the same inflammatory gene expression patterns as LPS in PBMCs of controls and PwD2. LEC could not upregulate the gene expression of IFN-γ, PSMB8, CD80, TCR, and MHC II involved in adaptive immune response and T-cell killing to the same extent in PBMCs of PwD2, as compared to non-diabetic controls. These data could explain why PwD2 are more vulnerable to bacterial and viral diseases as compared to non-diabetic individuals. These data also provide support for the hypothesis that bacterial and plant toxins (LPS/bacteria/viruses and lectins from plants) would react differentially to elicit the gene upregulation of important mediators in PBMCs from non-diabetic controls and PwD2.
This study also revealed for the first time that bacterial (LPS) and plant toxins may cause robust inflammation in healthy controls and PwD2, leading to the downregulation of the gene expression of the insulin receptor (INSR). In addition, the gene expression of INSRR and INSM1 in toxin treated PBMCs from PwD2 was upregulated to a lesser extent than non-diabetics, which may lead to insulin resistance and problems with cell maturity in the host. LPS is known to circulate in the bloodstream, adipose tissue, and the liver, to induce insulin resistance in vivo, but this has not been previously shown with LEC [70,71]. This study reports that PBMCs from people of the same age and weight respond differentially to LPS and LEC, depending on whether they were non-diabetic or PwD2. LPS and LEC also robustly activated the thyroid hormone receptor-interacting protein 10 (TRIP10, which can change the metabolic response) in both PBMCs, from nondiabetics and PwD2. The enzymes responsible for the biosynthesis of thyroid hormones are degraded by the proteasomes. Proteasome modulators RES inhibits LMP7 and reduces inflammation, while LEC (activates LMP7 and induces inflammation). Recently, we have shown that a mixture of RES, vitamin D3 and δ-tocotrienols (NS-3) as food supplements reduce blood sugar and inflammatory cytokines in PwD2 in a clinical study in vivo [72]. More genomic, proteomic research, and clinical studies using proteasome modulators in humans are, therefore, required to fully understand how various dietary components, such as lectins/flavonoids [73], affect lectin receptors, transcription factors, and proteasomes in DM2.

Experiments Prior to RNAseq Analysis
To gain a comprehensive picture for the contribution of dietary nutrients we had a total of n = 5 non-diabetic controls and n = 5 PwD2 (Table 3). The first four sets were analyzed using selected primers by RT-PCR to find the proper dose of modulators. The fifth sample with 16 treatments was analyzed by Novogene using next generation sequencing (NGS) using the Illumina supports, as described below and in reference [54].

Detection of Cell Viability and Isolation of Total Cellular RNA
The viability and number of PBMCs were determined by the trypan blue dye exclusion test and by using a cell counter. After treatment, cells were washed with PBS and total RNA was extracted by using RNeasy mini kit (Qiagen, Germantown, MD, USA) as per the manufacturer's instructions.

Total Cellular RNA Isolation and qPCR
qPCR was performed with total RNA isolated from cells treated with RES, LEC and/or LPS, as described previously [28,29,54]. Total cellular RNA was isolated with a RNeasy mini kit according to the manufacturer's instructions (RNeasy, Qiagen, Chatsworth, CA, USA). To check the purity of the RNA for cytokines, reverse transcription and PCR were conducted using one step qPCR Kit (Qiagen, Chatsworth, CA), according to the manufacturer's instruction.

Sample Preparation for RNAseq Analysis
Five to 8 µg of total RNA from each sample were provided to Novogene Global (Sacramento, CA, USA) for RNAseq analyses using the human Illumina system (Santa Clara, CA, USA), as described previously [54]. The purity of total RNA was assessed using the following tests, Nanodrop (OD 260/280), agarose gel electrophoresis and Agilent 2100 analysis to check RNA integrity. The following procedures were carried out by Novogene: mRNA enrichment, conversion to double-stranded cDNA, end repair, poly-A adaptor addition, fragment selection and PCR, library quality assessment, and Illumina sequencing. An average of 44-61 million raw read counts were obtained. Novogene used the STAR software for alignment for RNA-seq data analysis. Read counts are proportional to gene expression levels, gene length, and sequencing depth. The differential gene analysis was carried out on two samples (control vs. treatment group) at a time using the DESeq2 R package. The threshold of differential expression genes is log2 fold change >1, and p value < 0.05 [54].

Data Analysis, Network and Pathway Analysis
Gene expression data were first imported in the differentially expressed genes (DEG) program and numbers were corrected for differences in the IIlumina analysis. The numbers presented in Tables S1-S6 are averages from 2 incubations for each treatment; the log ratio values were normalized to a scale of 0 (instead of 1, which shows decimals), the expression values of upregulated genes showed positive numbers, and the downregulated ones showed negative numbers (called normalized ratios, a log ratio of 2 is equivalent to a fold change of 4); these ratios were imported into the ingenuity pathways analysis (IPA) software (Ingenuity Systems, Mountain View, CA, USA) [28,29,54]. This is a web-based tool that is predicated on more than 200,000 full-text articles and has information based upon 7900 human and mouse genes. This system categorizes genes into high-level cellular functions and canonical pathways and has been used to characterize genes important in human systemic inflammation [28,29,54]. Genes found to be significantly activated were categorized based on different pathways and networks available in the database and ranked by score, as described previously [28,29,54]. Genes identified as statistically different from the background, in terms of activation relative to control cells, were analyzed and mapped into different pathways.

Statistical Analysis
The data were analyzed using analysis of covariance (ANOVA) to compare means of pre-treatment versus post-treatment. Data are reported as mean + SD (standard deviation in Figure 8. The statistical significance level was set at 5% (p < 0.05) [54].

Conclusions
These findings provide insights into immunologic mechanisms operative in PwD2 in the presence of toxins. Collectively, these novel findings strongly support the existence of a highly dysregulated level of immune activity in PBMCs from PwD2 relative to nondiabetic adults. In PBMCs from non-diabetics, LPS induces gene expression of several cytokines. LPS induces gene expression of IL-12A in monocytes and macrophages, leading to the activation of IFN-γ from NK or Th1 cells. Th1 cells and dendritic cells signal via the interaction of CD80 with CD28, and MHC II molecules with T-cell receptor. IFN-γ amplifies the entire signaling pathway by upregulating LMP7 (PSMB8), LMP2 (PSMB9) and several induced proteins in PBMCs. The continued interaction of cells with LPS and IFN-g leads to tolerance, where there is no significant change in the gene expression of PSMB8 or other subunits of the proteasome with agonists. However, in PBMCs from PwD2 this regulation is lost because the cells do not robustly induce the gene expression of PSMB8 with LPS, and there is a weak interaction between the dendritic cells and Th1 cells. Therefore, induction of gene expression of IL-12A, IFN-γ, STAT1, MHC-II molecules, LMP7, and several IFN-γ-induced genes was not robustly elicited in response to LPS in PBMCs from PwD2, as compared to those from non-diabetics ( Figure 10). RES (with/without LPS) downregulates the gene expression of PSMB8 and the mediators of IFN-γ pathway even further, while LEC upregulates the gene expression of these mediators. The data presented in this manuscript would be very useful in designing a therapeutic approach for most inflammatory and infectious diseases, such as diabetes, sepsis, tuberculosis, glaucoma, macular degeneration, neuroinflammatory disorders, COVID-19, and AIDS. discussions throughout this study. We thank Neerupma Silswal (Kansas University) for obtaining data shown in Figure 8.