Replacement substance P reduces cardiac fibrosis in monkeys with type 2 diabetes

Abstract

Background:

Type 2 diabetes mellitus (T2DM)-associated cardiac fibrosis contributes to heart failure. We previously showed that diabetic mice with cardiomyopathy, including cardiac fibrosis, exhibit low levels of the neuropeptide substance P; exogenous replacement of substance P reversed cardiac fibrosis, independent of body weight, blood glucose and blood pressure. We sought to elucidate the effectiveness and safety of replacement substance P to ameliorate or reverse cardiac fibrosis in type 2 diabetic monkeys.

Methods:

Four female T2DM African Green monkeys receive substance P (0.5 mg/Kg/day S.Q. injection) for 8 weeks. We obtained cardiac magnetic resonance imaging and blood samples to assess left ventricular function and fibrosis by T1 map-derived extracellular volume as well as circulating procollagen type I C-terminal propeptide. Hematological parameters for toxicities were also assessed in these monkeys and compared with three female T2DM monkeys receiving saline S.Q. as a safety comparison group.

Results:

Diabetic monkeys receiving replacement substance P exhibited a ~20% decrease in extracellular volume (p = 0.01), concomitant with ~25% decrease procollagen type I C-terminal propeptide levels (p = 0.008). Left ventricular ejection fraction was unchanged with substance P (p = 0.42); however, circumferential strain was improved (p < 0.01). Complete blood counts, glycosylated hemoglobin A1c, lipids, liver and pancreatic enzymes, and inflammation markers were unchanged (p > 0.05).

Conclusions:

Replacement substance P reversed cardiac fibrosis in a large preclinical model of type 2 diabetes, independent of glycemic control. No hematological or organ-related toxicity was associated with replacement substance P. These results strongly support a potential application for replacement substance P as safe therapy for diabetic cardiac fibrosis.

1. Background

Patients with type 2 diabetes mellitus (T2DM) exhibit a higher risk of developing left ventricular (LV) dysfunction and heart failure. A subset of these patients develop diabetic cardiomyopathy independent of coronary atherosclerosis, hypertension or valvular disease [1–3]. Cardinal characteristics of this distinct cardiac entity include the presence LV hypertrophy and interstitial cardiac fibrosis [4], which results in increased cardiac stiffness that contributes to LV diastolic dysfunction [4,5]. Importantly, diffuse cardiac fibrosis determined by cardiac magnetic resonance (CMR)-derived extracellular volume (ECV) is an independent predictor of morbidity and mortality among diabetics [6,7] and yet, it remains one of the most challenging phenotypes to treat due to a lack of effective therapies.

We have recently reported that T2DM mice exhibit low circulating levels of the neuropeptide substance P (SP)[8], similarly to diabetic rats [9] and even T2DM humans [10]. It has been suggested that this loss of SP is due to diabetic neuropathy-associated damage of nerve fibers that release SP [10,11]. Remarkably, we were able to show in mice that exogenous replacement of SP could reduce interstitial cardiac fibrosis, independent of body weight, blood glucose, and blood pressure [8]. The mechanisms underlying the anti-fibrotic actions of replacement SP were the ability of SP to oppose the pro-fibrotic cardiac fibroblast phenotype induced by high glucose, as well as promoting a shift from the pro-inflammatory M1 macrophage phenotype to the M2 anti-inflammatory/reparative phenotype.

As a result of these intriguing finding in a murine model of T2DM, we wanted to explore the potential translational significance by assessing whether replacement SP is an effective and safe treatment strategy to mitigate T2DM-induced cardiac fibrosis in an African Green monkey model of spontaneous T2DM. This large animal model closely recapitulates human diabetic cardiomyopathy and exhibits low circulating levels of SP compared to age-matched controls. The primary objectives of this study were 1) to quantify CMR-derived ECV and cardiac function changes before and after replacement SP and 2) determine if replacement SP is a feasible and safe treatment strategy.

2. Methods

2.1. Animals

The study subjects were sourced from a multigenerational pedigreed colony of African Green monkeys (Chlorocebus aethiops sabeus; n = 311, 4–27 years, lifespan ≈ 26 years), which descended from 57 founder monkeys at the Wake Forest Vervet Research Colony [12]. Four T2DM monkeys (ages 14 – 21 years) that had been diagnosed at least 12 months prior were treated with replacement SP (0.5 mg/Kg/day). Additional 3 animals (one T2DM and two insulin resistant [IR] monkeys) treated with saline were used as safety comparison group (Fig. 1). All procedures were performed in accordance with the Guide for Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee (A19–006) at our Association for the Assessment and Accreditation of Laboratory Animal Care International accredited institution, which operates in compliance with the Animal Welfare Act. T2DM monkeys were given twice-daily insulin therapy to avoid ketoacidosis, and insulin therapy had been stable for at least 6 months prior to study initiation. To scale the dose of SP successfully used in our previous mouse study (1 mg/Kg/day) to monkeys, we conducted an allometric scaling for dose conversion that accounts for the difference in physiological time between the species [13]. Animals were fed a commercial laboratory primate chow (Laboratory Diet 5038; LabDiet, St. Louis, MO) comprised of 13% of calories from fat, 69% of calories from carbohydrates, and 18% of calories from protein, with daily supplemental fresh fruits and vegetables and tap water ad libitum. Monkeys were socially housed in a climate-controlled environment.

Fig. 1.

Study Design. Type 2 diabetic African Green monkeys received daily replacement of substance P (n = 4) for 8 weeks. Baseline and post-treatment cardiovascular magnetic resonance (CMR) and adipose tissue biopsies were performed 2 weeks before the initiation and during the last two weeks of treatment. Blood and urine were collected before (baseline), 2 (MP-1), and 4 weeks (MP-2) after the first injection and at the end of treatment (post-treatment, Post-Tx). The safety comparison group received vehicle (n = 3) for 8 weeks and underwent the same procedures at the same time points.

2.2. Experimental design and procedures

Two weeks prior to treatment initiation, animals underwent procedures that included CMR, blood pressure measurement, blood and urine collection, and pulse wave velocity (PWV) assessment and subcutaneous fat biopsies (Fig. 1). Following study initiation, animals were treated with SP for 8 weeks. Blood and urine collections were again performed at weeks 2 (MP-1) and 4 (MP-2) during the treatment period to monitor hematological parameters, comprehensive metabolic biochemistry panels and circulating biomarkers. At completion of the 8-week study period, CMR, blood pressure measurement, blood and urine collection, and PWV assessment and subcutaneous fat biopsies were repeated. The safety comparison group underwent the same procedures.

Procedures:

Animals were fasted overnight and anesthetized with intramuscular ketamine (10–15 mg/kg) and midazolam (0.05 mg/kg) for imaging, blood and urine, and morphometric data collections. Triplicate measures of blood pressure (systolic and diastolic blood pressure [SBP and DBP]) were taken indirectly by high definition oscillometry at the tail base, and the average value reported. PWV was calculated in triplicate using the femoral and brachial arteries for pulse measurement sites and distances (SphygmoCor, AtCor, Naperville, IL). Blood samples were obtained by venipuncture of the femoral vein and collected into ethylenediaminetetraacetic acid (EDTA) and serum separator blood tubes. Urine samples were obtained by suprapubic bladder aspiration. After processing, the plasma, serum and urine samples were stored at − 80 °C until analysis. Adipose biopsies were collected from the para-umbilical region. A portion of the samples were fixed in 4% paraformaldehyde for histopathology and immunohistochemistry.

2.3. Cardiac magnetic resonance imaging

Animals underwent CMR to assess cardiac structure and function using a 3.0 T Siemens Skyra scanner (Siemens Medical Systems, Erlangen, Germany). ECV was calculated from T1-maps acquired pre-contrast (native T1) and 12 min post-contrast administration. Offline image analysis was performed in a paired, blinded analysis by an experienced observer (GCM), using software specifically designed for this study. LV end-diastolic volume (LVEDV), end-systolic volume (LVESV), stroke volume, myocardial mass and global circumferential strain were determined from cine white blood steady-state free precession (SSFP) images (192 ×128 matrix, 20-cm field of view, 10-ms repetition time (TR), as 4-ms echo time (TE), a 35-degree flip angle, a slice 6-mm thick, and a 40-ms temporal resolution). Cine function and strain analysis was performed off-line and in a paired, blinded fashion by an experienced imager (GCM) using commercial software (Segment CMR, MEDVISO). Endo- and epicardial contours were traced semi-automatically for LV volume calculations at end diastole and end systole on off-line workstations and summed using Simpson’s rule for calculation of LV ejection fraction (EF) as previously described [14,15]. Feature tracking segmentation was performed in short-axis views of the base, mid and apex.

2.4. T1 mapping and extracellular and total cell volumes

Glomerular filtration rates (assessed by serum creatinine levels) were measured in all animals from venous blood samples drawn within a week prior to the CMR exam. Wake Forest Primate Center veterinarian (KK) confirmed intact kidney function and authorized the administration of 0.15 mmol/kg of gadolinium contrast (Prohance, Bracco Diagnostics, Princeton, New Jersey). Using a modified Look-Locker inversion recovery sequence (MOLLI), T1 mapping was performed in a mid-cavity short-axis slice [16]. The MOLLI imaging acquired 11 images in the span of 17 heartbeats with 220 × 180 mm field of view collected with a 192 × 128 matrix, 35 flip angle, 6 mm slice thickness, 1.1 ms echo time, 2.2 ms repetition time, and generalized auto-calibrating partially parallel acquisition factor of 2 [17]. As previously published, a 3-parameter curve fit was applied to MOLLI source images to create T1 maps [18]. Endocardial and epicardial borders were traced manually on each T1 map to ensure the LV blood pool and epicardial fat were excluded. The pre- and post-contrast maps were examined in parallel to avoid resolution issues derived from measurements performed directly on ECV maps. Then, the myocardium was segmented using a spoke tool to divide the LV into the 6 mid-cavity segments according to the American Heart Association standardized method [19]. Each segment of the mid LV short axis was assessed for image artifacts, delineation of myocardial blood pool borders and homogeneity of myocardial signal. For reliable ECV calculations, segments that exhibited artifacts, inhomogeneity or thin stripe of myocardium less than 2 pixels thick were excluded from the analysis [16]. The average T1 for each map both pre- and post-contrast, was calculated from accepted segments only. For the ECV fraction calculation, an additional region of interest was placed in the LV blood pool to determine the T1 of blood in each map. Hematocrit laboratory values obtained the same day of the image acquisition were used to determine the following parameters: a) ECV was calculated using pre- contrast and post-contrast values and the partition coefficient was determined from the slope of the line of the variables 1/T1myo versus 1/T1blood as previously described [16,20,21]; b) intracellular volume (ICV), measured with the formula ICV = 1 − ECV and c) total cell volume was calculated by the formula Total Cell Volume = ICV × LV myocardial volume indexed as previously described [22]; LV volume was calculated by normalizing LV mass to 1.06.

2.5. Hematologic and metabolic monitoring

To evaluate safety of replacement SP therapy and to identify potential side effects, we assessed complete blood counts (CBC), comprehensive metabolic panel, lipid panel (cholesterols), liver enzymes (total bilirubin, alkaline phosphatase [ALP], aspartate aminotransferase [AST] and alanine aminotransferase [ALT]), pancreatic function (lipase, amylase), and kidney function (creatinine and blood urea nitrogen). All clinical laboratories were processed by IDEXX (Greensboro, NC, USA).

2.6. Circulating biomarkers

Blood biomarkers (N-terminal pro b-type natriuretic peptide [NT-pro-BNP], procollagen I C-terminal propeptide [PICP], tumor necrosis factor [TNF]-α, interleukin [IL]–6, soluble suppression of tumorigenicity 2 [ST2], IL-33, matrix metalloproteinase [MMP]9, MMP1) were measured using commercially available ELISA kits optimized for monkey samples (MyBiosource, San Diego, CA). SP was measured with a commercially available ELISA kit (R & D Systems, Minneapolis, MN) according to manufacturer’s instructions, including addition of aprotinin (Tocris^™, Catalog # 4139) that was added within 5 min of the serum collection to stabilize the neuropeptide. Biomarkers were measured in duplicate and reported as the mean value.

2.7. Macrophage labeling and phenotyping

Subcutaneous adipose tissue biopsies from the para-umbilical region were embedded in paraffin blocks and cut into 5 μm sections. Adipose tissue was selected due to ease of access with the assumption these cells are exposed to comparable SP concentrations as the myocardium. Further there is a close connection between the inflammatory state observed in T2DM and obesity, both known to negatively influence cardiovascular health. These sections were immunofluorescently labeled by StageBio (Mt Jackson, VA) to identify macrophage phenotypes as follows: M1 = CD68⁺CD163⁺pSTAT⁺; M2 = CD68⁺CD163⁺cMAF⁺; and Mϕ undifferentiated = CD68⁺CD163⁺pSTAT⁻cMAF⁻. DAPI was used to identify cell nuclei. Cell numbers and tissue area were determined using a customized digital image analysis application (Visiopharm, Hoersholm, Denmark).

2.8. Statistical analysis

All grouped data were expressed as mean ± SD. Effectiveness of replacement SP was tested by comparing the baseline and post-SP parameters by paired t-tests. For the evaluation of safety, the replacement SP and safety comparison groups were compared by two-way ANOVAs with post-hoc Tukey or Bonferroni tests as appropriate. Statistical significance was determined when p < 0.05. All analysis was performed using GraphPad Prism version 9 (San Diego, CA).

3. Results

3.1. Biometric parameters

Biometric parameters for the replacement SP trial are displayed in Table 1. Replacement SP induced a significant post-treatment reduction in body weight of the monkeys (p = 0.02). There were no statistical differences between baseline vs. post-treatment in age, PWV, SBP, DBP, HbA1c (p > 0.05).

Table 1.

Biometric Parameters.

Characteristic Mean ± SD	Safety Comparison Group			Replacement SP Group
	Baseline	Post-Tx	p-value	Baseline	Post-Tx	p-value
Age, yrs	18 ± 3.5			21 ± 0.8
Body Weight, Kg	6.3 ± 2.0	6.0 ± 1.9	0.03	7.1 ± 1.6	6.5 ± 1.4	0.02
BSA, m 2	0.3 ± 0.06	0.3 ± 0.06	0.03	0.3 ± 0.04	0.3 ± 0.04	0.02
PWV	9.2 ± 1.6	7.2 ± 2.5	0.24	9.4 ± 2.4	6.8 ± 0.9	0.07
Blood Pressure, mmHg
Systolic	124 ± 27.8	129.6 ± 38.9	0.27	119.8 ± 35.5	125 ± 8.8	0.40
Diastolic	77.1 ± 21.9	71.7 ± 19.5	0.04	74.42 ± 16.9	69.42 ± 5.4	0.25
Metabolic
HbA1C, mmol/mol	7.3 ± 1.6	6.7 ± 1.1	0.13	8.6 ± 1.6	8.1 ± 1.4	0.11
Glucose, mg/dL	125 ± 54.6	132 ± 15	0.4	255 ± 88.5	331.2 ± 92.9	0.04

SP: substance P; SD: standard deviation; Post-Tx: post-treatment; BSA: body surface area; PWV: pulse wave velocity

3.2. Replacement SP Reduces CMR and Circulating Markers of Cardiac Fibrosis

Monkeys exhibited an average of 20.1% decrease in CMR-derived ECV from post-replacement SP (Fig. 2A, and supplemental Fig. 5), total cell volume did not change significantly from baseline (Fig. 2B, p = 0.21) and circulating PICP was significantly decreased post-replacement SP experimental period (Fig. 2C, p = 0.008, and Supplemental Fig. 4).

Fig. 2.

Replacement substance P reduces CMR-derived extracellular volume and circulating markers of cardiac fibrosis in T2DM monkeys. Box and whiskers plots of A) Extracellular Volume (ECV), B) Total Cell Volume, C) circulating procollagen type I C-terminal propeptide (PICP) analyses, mean ± min and max data at baseline (white boxes) to post SP treatment (Post SP, blue boxes) in T2DM monkeys receiving the replacement substance P; and D) representative ECV maps from T2DM monkeys in the replacement SP.

3.3. Cardiac structure and LVEF was unchanged but myocardial strain improved by replacement SP

CMR-derived structural and functional parameters are displayed in Table 2. There were no statistically significant changes in LVEF or LV myocardial mass indexed to body surface area (BSA) between baseline and post-treatment (Table 2, Fig. 3 and Supplemental Fig. 6). Despite the fact that SP did not impact global cardiac function, circumferential strain measurements showed an improvement compared to baseline at the basal (p = 0.005), mid ventricular (p = 0.01) and apical level (p = 0.01) (Fig. 3C).

Table 2.

Cardiovascular magnetic resonance imaging measures in the replacement SP group.

Characteristic Mean ± SD	Replacement SP Group
	Baseline	Post-Tx	p-value
HR, bpm	116 ± 20.9	97.3 ± 6.1	0.06
LVEF, %	73.2 ± 10.6	71.7 ± 4.6	0.4
LVEDV, ml	8.3 ± 0.5	6.8 ± 2.06	0.1
LVEDVi, ml/m 2	24.6 ± 3.2	21.03 ± 6.6	0.3
LVESV, ml	2.3 ± 0.9	1.8 ± 0.9	0.3
LVESVi, ml/m 2	6.5 ± 2.6	5.6 ± 3.2	0.4
LV mass, g	9 ± 3.4	8.8 ± 2.1	0.4
LV massi, g/m 2	26 ± 6.1	27.1 ± 5.4	0.4
Native T1, ms	1424.8 ± 152.9	1477 ± 230.8	0.3
T1 Post-contrast, ms	401.5 ± 9.3	434.8 ± 55.8	0.2
IVC	0.6 ± 0.05	0.7 ± 0.03	0.015

SP: substance P; Post-Tx: post-treatment; HR: heart rate LVEF: left ventricular ejection fraction; LVEDV: left ventricular end-diastolic volume; LVEDVi: left ventricular end-diastolic volume index; LVESV: left ventricular end-systolic volumes; LVESVi: left ventricular end-systolic volumes index; LV massi: left ventricular mass index; IVC: intracellular volume

Fig. 3.

Replacement substance P improves myocardial strain, but it did not influence left ventricular structure. A) Left ventricular ejection fraction (LVEF), B) left ventricular mass indexed to body surface area and C) circumferential strain measurements at baseline (white bars) and post-SP (blue bars) at basal, mid-ventricular and apical levels in T2DM monkeys receiving the replacement substance P.

3.4. Assessment of circulating markers

NT-pro-BNP progressively decreased with replacement SP and there was a statistically significant difference between MP-1 and post-SP (p = 0.04, Fig. 4A and Supplemental Figure 7 A). The baseline levels of NT-pro-BNP were significantly different (p = 0.01, Fig. 4A) and there were no changes among time point in the safety comparison group. There were no significant differences for TNF-α or IL-6 (Fig. 4B and C). sST2 progressively decreased in the safety comparison group and there was a statistically significant decrease from baseline at MP-2 (p = 0.035, Fig. 4D, and Supplemental Figure 7D); this difference was gone by the final time point. There were no intra- or intergroup differences for IL-33, MMP-9, or MMP-1 (Fig. 4E–G and Supplemental Figure 7E–G).

Fig. 4.

Replacement SP improved cardiac, but not inflammatory or other biomarkers in T2DM monkeys. A) N-terminal-pro B-type natriuretic peptide (NT-pro-BNP), B) Tumor necrosis factor (TNF)-α, C) interleukin (IL)– 6, D) soluble suppression of tumorigenicity 2 (sST2), E) IL-33, F) matrix metalloproteinase (MMP)– 9 and G) MMP-2 in monkeys in the replacement substance P (SP, blue lines) or safety comparison (red lines) groups. Values are reported as mean ± SD.

3.5. Adipose tissue macrophage profiles showed a shift towards an undifferentiated phenotype with replacement SP

Replacement SP did not alter the M1 (CD68 +CD163 +pSTAT+), M2 (CD68 +CD163 +cMAF+) macrophage density or cells labeling as both M1 and M2 macrophages (CD68⁺CD163⁺pSTAT⁺ cMAF⁺) (p > 0.05; Fig. 5A–C). There was, however, a significant increase in undifferentiated macrophages in the replacement SP group (p = 0.04, Fig. 5D). This resulted in a time-dependent shift in differentiated (M1/M2) versus undifferentiated macrophage profiles, with replacement SP promoting a shift towards more macrophages of an undifferentiated phenotype (Fig. 5E).

Fig. 5.

Substance P promoted a shift from differentiated macrophages to undifferentiated macrophages in adipose tissue. Adipose macrophage density (number of macrophages per cm²) in T2DM monkeys after replacement substance P (blue boxes) compared to baseline (white boxes) A) M1 macrophages: CD68⁺CD163⁺pSTAT⁺; B) M2 macrophages CD68⁺CD163⁺cMAF⁺; C) M1/M2 macrophages: CD68⁺CD163⁺pSTAT⁺cMAF⁺; D) Undifferentiated macrophages CD68⁺CD163⁺pSTAT⁻cMAF⁻; E) Pie charts showing the proportion of undifferentiated macrophages (CD68⁺CD163⁺pSTAT⁻cMAF⁻, red) and differentiated macrophages (M1 [pSTAT⁺], M2 [cMAF⁺] and M1/M2 [pSTAT⁺ and cMAF⁺] macrophages); and F) Representative Immunoflourescent images of adipose tissue; G) undifferentiated macrophage (red arrow), M1-polarized macrophage (green arrow) and M2 macrophage (magenta arrow) are indicated on the zoomed in image panel. Each monkey had comparable total tissue macrophage density (0.12/mm²); differences in undifferentiated macrophages (red) can be appreciated.

3.6. Replacement SP is Safe and Tolerated by Monkeys

There were no differences in laboratory or metabolic parameters at baseline between the groups (p > 0.05 for all, Supplemental Tables 4 and 5). Replacement SP did not have a clinically meaningful effect on blood glucose (Table 1, Supplemental Table 4) or HbA1c (ΔA1c, SP: −0.45 ± 0.59% vs vehicle: −0.53 ± 0.61%, p = 0.43, Table 1, Supplemental Table 4). Complete blood counts, metabolic panel including lipids, liver, kidney, and pancreatic enzymes were unchanged and within normal clinical ranges (p > 0.05 for all; Fig. 6, Supplemental Figure 8, and Supplemental tables 4 and 5). Based on these data we concluded that the SP therapy was safe at the dose and duration tested.

Fig. 6.

Impact of replacement SP in liver, pancreatic and kidney function. Raw values of A) Total bilirubin, B) alkaline phosphatase (ALP), C) aspartate aminotransferase (AST), D) alanine aminotransferase (ALT), E) amylase, F) lipase and G) BUN/Creatinine ratio in monkeys in the replacement substance P (SP, blue lines) or the safety comparison group (red lines) groups. Values are reported as mean ± SD. The time-dependent changes in ALT occurred in all animals and they are likely due to repeated sedations. These ALT effects have been previously observed in monkeys and they are not the result of the intervention.

4. Discussion

T2DM-associated cardiac fibrosis continues to be a critical contributor to the development of LV diastolic dysfunction and heart failure. Pharmacological strategies that inhibit progression and reverse cardiac fibrosis remain elusive. Recently, we identified that the loss of SP that occurs in diabetes contributes to the establishment of cardiac fibrosis and that exogenous replacement of SP was an effective anti-fibrotic strategy in mice[8]. Despite the identification of many potential targets for treating fibrosis using small animal models, there are still no effective treatment strategies to ameliorate or reverse fibrosis. We sought to overcome this limitation by validating our mouse findings in an African Green monkey model of T2DM. This translationally relevant model displayed cardiac fibrosis in response to T2DM (Supplemental Fig. 1), similar to T2DM rhesus monkeys [23], and to the clinical condition in humans [24]. Importantly, this monkey model also exhibited a loss of SP (Supplemental Fig. 3B). Their size and unique genetic traits shared with humans [25], the ability to use clinical cardiac imaging techniques, and similarities in the clinical manifestation of T2DM [26], means that this monkey model represents a valuable intermediate step between rodents and patients to assess the effectiveness and safety of the therapeutic strategy. In this regard, we demonstrate that replacement SP 1) decreased cardiac fibrosis, 2) improved LV global circumferential strain, and 3) was not associated with organ-related toxicities and did not affect glycemic control or insulin requirements. These results strongly support replacement SP as a novel, effective, and safe anti-fibrotic therapeutic strategy for T2DM-induced cardiac fibrosis.

In this preclinical monkey trial, we used female T2DM monkeys with an age range between 14 and 19 years. Diabetic monkeys receiving replacement SP exhibited a decrease of ~20% in ECV after treatment (Fig. 2A). While ECV has been traditionally recognized as a CMR-derived metric that quantifies diffuse myocardial fibrosis [24], it is important to recognize that ECV is the ratio of the volume of the interstitial space and the total volume of myocardial tissue. As such, the mechanisms by which ECV decreases can be classified by factors that a) decrease the interstitial space such as a decrease of extracellular matrix or b) increase in the total cell volume of myocardium, mainly afforded by increased cardiomyocyte size [22,27,28]. Because diabetes induced-cardiac fibrosis typically occurs concomitantly with cardiomyocyte hypertrophy [29, 30], it was important to assess whether replacement SP affected cardiomyocyte size. We found that none of animals receiving SP experienced significant changes in the total cell volume (Fig. 2B) or LV mass (Table 2). Additionally, the ECV decrease in the SP group was similar to the ~25% decrease of circulating PICP, a marker of cardiac fibrosis [31, 32] (Fig. 2C). Therefore, ECV decreases observed in our study are likely afforded by a decrease in interstitial cardiac fibrosis. These findings are consistent with the histopathological amelioration of cardiac fibrosis and no changes in cardiomyocyte size in diabetic mice that received replacement SP in our previous study [8]. This is also in agreement with a study that found no cardiac hypertrophy in diabetic monkeys [23].

In our study, replacement SP did not have an impact on LVEF (Fig. 3A). However, despite a normal LVEF, global myocardial strain significantly improved with replacement SP (Fig. 3C). Previous studies have shown that speckle-tracking strain impairment is an early marker of diabetic cardiomyopathy in T2DM patients with normal systolic function [33,34]. It is plausible that the improvement in cardiac strain in our monkeys may be associated with the amelioration of cardiac fibrosis. Prior studies have demonstrated that both cardiac hypertrophy and fibrosis influence CMR-derived circumferential strain deformation [35]. While disarrangements of the cardiomyocyte compartment are the main factors driving circumferential strain impairments [36] and cardiac fibrosis is only moderately associated with cardiac deformation[37], we speculate that the modest improvement in circumferential strain in our animals may be due to the decrease in collagen fibers tethered to cardiomyocytes. While there are no studies assessing the direct effects of SP on cardiomyocyte contractility, there are studies that have shown that SP can cause smooth muscle contraction, likely through extracellular calcium influx via voltage-operated and receptor-operated calcium channels [38–40]. Thus, we cannot rule out direct effects of SP on cardiomyocytes to improve contractility. Interestingly, while we observed reduced fibrosis and improvements in cardiac contractility with replacement SP, NT-pro-BNP - a predictor of cardiovascular events and mortality among T2DM patients [41,42]- was also significantly reduced post-replacement SP. This may be indicative of improvements in hemodynamic load on the heart due to SP-induced reversal of cardiac remodeling.

Another important underlying mechanism that may have contributed to the improvement of cardiac strain is SP mitigation of oxidative stress induced by high glucose [43]. Our laboratory and others have shown that high glucose promotes the production of reactive oxygen species by cardiac fibroblast [8] and endothelial cells [8,44] by increasing formation of advanced glycation products (AGE) and the activation of their receptor (RAGE). In our recent article in mice, we reported that in cardiac fibroblasts exposed to high glucose levels, SP decreases RAGE and oxidative stress [8]. Interestingly, this was accompanied by an increase in nitric oxide (NO) production that could also account for the potential beneficial effects of SP. A similar finding was observed in cardiac microvasculature endothelial cells under hyperglycemic conditions, where SP also restored NO production [44].

Interestingly, our recent mouse study demonstrated that replacement SP prevented the influx of pro-inflammatory M1 macrophages into the diabetic heart, and potentially increased anti-inflammatory/reparative M2 macrophages to ultimately cause a shift to more favorable M2 to M1 ratio [8]. The M1 macrophage phenotype is characterized by production of TNF-α and IL-6, among other molecules. Consequently, it was somewhat surprising that we did not find reduced TNF-α or IL-6 in response to replacement SP in the monkeys. There are at least two potential explanations for this discrepancy. The first is that we measured circulating levels of TNF-α and IL-6 and this may not be reflective of these cytokines locally in the heart. The second potential explanation relates to complexities in macrophage phenotype in primates versus rodents. In diabetic and obese rodents, there is an influx of Ly6C⁺ monocytes into numerous tissues, including the heart [8,45] that then differentiate into M1 macrophages, leading to an imbalance between M1 and M2 phenotypes [46,47]. In rodents, there is a clear functional difference with M1 macrophages being pro-inflammatory and producing cytokines such as TNF-α, IL-6 and CCL2, and M2 macrophages being anti-inflammatory/reparative and producing markers including arginase-1 and IL-10 [48]. While in human (primate) diabetes and obesity, adipose macrophages also appear to be of both the M1 and M2 phenotype [49], the differences in M1 and M2 function do not appear to be as clear-cut. Recent evidence suggests that in human diabetes and obesity, adipose macrophages are inflammatory in nature regardless of M1 and M2 status[49]. This may well be what is reflected in the macrophage data in our current study. In our monkey study, we labeled M1 (CD68⁺CD163⁺pSTAT⁺) and M2 (CD68⁺CD163⁺cMAF⁺) macrophages in adipose tissue as a surrogate for macrophages in the heart. We found no difference between the baseline and post-SP treatment in terms of M1 or M2 phenotypes. This is in total contrast to what we observed in the mouse heart, where replacement SP dramatically reduced M1 macrophages [8]. Interestingly though, in the monkeys there was a shift towards more macrophages of an undifferentiated phenotype (CD68⁺CD163⁺pSTAT⁻cMAF⁻) with replacement SP, which we would argue indicates a shift away from an activated phenotype (M1/M2) to a more quiescence, inactive phenotype (Fig. 5). Therefore, in essence, achieving the same result as the shift in balance to a more favorable M2 to M1 ratio that we observed in the hearts of mice receiving replacement SP. This likely speaks to the complexity of macrophage phenotype as well as important differences between rodent and primate macrophages. Regardless, we observed a clear reduction in fibrosis in T2DM monkeys (ECV, PICP) in response to replacement SP that could potentially be due, at least in part, to the shift towards the inactive macrophage phenotype.

An important objective of our study was to evaluate the feasibility and safety of exogenous replacement SP. SP is a neuropeptide involved in pain transmission, a mast cell degranulator [50,51] and activator of inflammatory cells that result in inflammation, vasodilation and plasma extravasation [52,53]. Despite this, the subcutaneous SP injections were well-tolerated, no allergic reactions or local pain, or nausea or vomiting were observed. Additionally, we performed complete blood count and chemistry panels, as well as assessment of liver, kidney and pancreatic function. No adverse effects of replacement SP were observed for any parameter (Fig. 6).

Together, the effectiveness of replacement SP to reverse cardiac fibrosis associated with T2DM and the absence of major side effects may represent a paradigm shift in therapeutic strategies to ameliorate T2DM-induced cardiac fibrosis in humans. Clinical studies have confirmed that a) circulating levels of SP are significantly decreased in Type 1 DM patients, especially those with neuropathy[54], and b) SP protein expression in atrial tissue of T2DM patients is also impaired and is associated with HF[10,55]. Larger studies confirming these findings in T2DM patients that account for comorbidities (hypertension, coronary artery atherosclerosis, prior myocardial infarction) are warranted. Moreover, further studies should investigate the effects of long-term replacement SP in combination with anti-diabetic and potentially anti-fibrotic medications (ie, sodium glucose cotransporter 2 [SGLT2] inhibitors, and glucagon-like peptide 1 [GLP-1] analogues) and validate these finding in human subjects.

The results of our study should be interpreted in the context of the following limitations: First, because our animals develop spontaneous T2DM, the number of monkeys available for this trial was limited. Second, it must be recognized that the safety-comparison group consisted of two IR monkeys and one T2DM monkey. However, we did not conduct a comparison of CMR-endpoints between the replacement SP group and the safety-comparison group due to the heterogeneity of cardiovascular burden. The fundamental purpose of having a safety comparison group was to determine whether the potential side effects of replacement SP could be due to pain or stress caused by daily injections or sedation for imaging, blood draws, and tissue biopsies. Evaluating safety and side effects of replacement SP was considered of high translational importance in this study. Additionally, the selection bias was allowed because the SP treated monkeys served as their own controls, as well as limited availability of T2DM animals. Third, diabetic safety comparison monkeys began the trial with significantly less cardiac fibrosis, compared to those in replacement SP group (p = 0.04, Table 2 and Supplemental Table 3) and the ECV baseline values of the cohort were highly variable (Supplemental Fig. 5). Nonetheless, decreases in ECV and strain improvement were only observed in replacement SP T2DM monkeys independent of their baseline ECV (Supplemental Fig. 5). Together, this highlights the translational value of the model as it reflects the heterogeneity of human condition[24] and the length of disease progression and glycemic control since T2DM diagnosis and emphasizes the effectiveness of replacement SP since this group contained the sicker animals.

5. Conclusions

In summary, herein we show in a large animal pre-clinical model of T2DM that replacement SP represents a novel, effective and safe adjuvant treatment strategy to reverse cardiac fibrosis, independent of glycemic control and anti-diabetic medications. This is in agreement with our previous mouse study[8] and therefore highlights the translational potential of this approach. Further studies should investigate the effects of long-term replacement SP in combination with anti-diabetic and potentially anti-fibrotic medications (ie, sodium glucose cotransporter 2 [SGLT2] inhibitors, and glucagon-like peptide 1 [GLP-1] analogues) and validate these finding in human subjects.

Abbreviations:

CCL2

C-C motif ligand 2

CMR

Cardiac magnetic resonance

CTL

Control

DBP

Diastolic blood pressure

BSA

Body Surface Area

ECV

Extracellular volume

IR

Insulin resistant

LV

Left ventricle

LVEDVi

Left ventricular end-diastolic volume index

LVEF

Left ventricular ejection fraction

LVESVi

Left ventricular end-systolic volume index

MMP

Matrix metalloproteinase

MOLLI

Modified Look-Locker inversion recovery sequence

Mϕ

Macrophage

NT-pro-BNP

N-terminal pro B-type natriuretic peptide

PICP

Procollagen type I C-terminal propeptide

PSR

Picrosirius red

PWV

Pulse wave velocity

SBP

Systolic blood pressure

SD

Standard deviation

SP

Substance P

ST2

Soluble suppression of tumorigenicity 2

T2DM

Type 2 Diabetes Mellitus

TNF-α

Tumor necrosis factor alpha

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.biopha.2023.114365.

Data Availability

Data will be made available on request.