Objectives: Posttransplant diabetes mellitus is a chronic metabolic complication that often develops in kidney transplant recipients and is an inflammatory disease that directly affects a patient’s immune system. Angiopoietin and angiopoietin-like proteins are intrinsic mediators of immune cells. We assessed the relationship between circulating angiopoietin proteins ANGPT¹ and ANGPT² and angiopoietin-like proteins ANGPTL³, ANGPTL⁴, ANGPTL⁶, ANGPTL⁷, and ANGPTL⁸ in kidney transplant recipients who develop posttransplant diabetes mellitus (experimental group) versus recipients who do not develop diabetes after transplant (control group).
Materials and Methods: We included 154 age-matched and sex-matched participants (38 with posttransplant diabetes mellitus) and 155 control participants. We collected 3 mL of venous blood from each patient. Plasma levels of angiopoietin proteins ANGPT¹ and ANGPT² and angiopoietin-like proteins ANGPTL³, ANGPTL⁴, ANGPTL⁶, ANGPTL⁷, and ANGPTL⁸ were determined by enzyme-linked immunosorbent assay. We determined the correlation between plasma levels of ANGPT¹ and ANGPT² and ANGPTL³, ANGPTL⁴, ANGPTL⁶, ANGPTL⁷, and ANGPTL⁸ in our posttransplant diabetes mellitus group and the control group.
Results: There were highly significant differences between kidney transplant recipients with diabetes versus control participants with regard to plasma levels of ANGPT¹, ANGPTL⁶, ANGPTL⁷, and ANGPTL⁸ (P < .⁰⁰¹, P = .⁰⁰¹, P < .⁰⁰¹, and P =.⁰⁴, respectively). No significant statistical association was found for ANGPT², ANGPTL³, and ANGPTL⁴ in our posttransplant diabetes mellitus cohorts.
Conclusions: Plasma levels of ANGPT¹, ANGPTL⁶, ANGPTL⁷, and ANGPTL⁸ may correlate with disease severity and chronicity in transplant patients who develop posttransplant diabetes mellitus. Thereby these may serve as potential biomarkers of the disease severity with possible early intervention in the management strategy especially with modifiable risk factors such as immunosuppression regimen, antidiabetic medications, and follow-up of the complications of posttransplant diabetes mellitus.

Key words : Enzyme-linked immunosorbent assay, Kidney transplantation

Introduction
Introducing new steroid-free immunosuppression protocols would significantly reduce many adverse side effects in kidney transplant recipients, including posttransplant diabetes mellitus (PTDM) risk. However, even without corticosteroids, the risk of PTDM remains a concern.¹ The oral glucose tolerance test has been shown to be essential for diagnosis and screening of cases with PTDM because alternative methods such as the hemoglobin A1c (HbA1c) test lack diagnostic sensitivity^2,3 and are associated with adverse outcomes.⁴ Importantly, the oral glucose tolerance test allows earlier identification of at-risk individuals on transplant wait lists.⁵ The HbA1c test, commonly used for diagnosing and monitoring diabetes, is unsuitable for patients on transplant wait lists or within the first 3 months after transplant. Anemia, which is common during this period, can distort the results of the HbA1c test. However, the HbA1c test becomes reliable for diagnosing and monitoring new-onset diabetes mellitus after transplant beyond 3 months after transplant, once blood counts normalize.⁶ Angiopoietins are growth factors involved in vascular repair and vasculogenesis. Angiopoietins ANGPT¹ and ANGPT² are major isoforms that regulate vascular homeostasis.^7-9 The ANGPT¹ isoform binds to its receptor tyrosine kinase with immunoglobulin, and epidermal growth factor homology domain-2 (TIE-2) stabilizes the vessel wall.⁷ In contrast, ANGPT², by either interacting with integrins or competing with ANGPT¹-TIE-2 receptor binding (inhibiting ANGPT¹-mediated TIE-2 phosphorylation), promotes vessel wall destabilization and favors, in the presence of vascular endothelial growth factor A, endothelial cell proliferation and new vessel formation.¹⁰ Angiopoietin-like proteins contain a coiled-coil domain and a fibrinogen-like domain, both characteristic of angiopoietins. Interestingly, angiopoietins also function in angiogenesis by regulating the survival and migration of endothelial cells, although angiopoietin-like proteins do not bind the angiopoietin receptor TIE-2. Presently, angiopoietin-like proteins are orphan ligands, but they have been reported to have pleiotropic effects on vascular cells, metabolism, and tumor biology.¹¹ Recently, the roles of angiopoietins and angiopoietin-like proteins in vascular biology molecular mechanisms relevant to these factors and angiogenesis have been discussed.¹² Kidney transplant is the treatment of choice for many people with severe chronic kidney disease because quality of life and survival (life expectancy) are often better than for patients treated with dialysis.^13,14 The introduction of new steroid-free immunosuppression regimens would significantly reduce many adverse side effects, including PTDM risk, but, in fact, even without corticosteroids, the risk of PTDM remains a concern. For example, obesity in kidney transplant candidates has been considered a well-established risk factor for PTDM.¹ Various functions of angiopoietin proteins and angiopoietin-like proteins have also been described in developmental, physiological, and pathophysiological processes.¹⁵ Crucially, the angiopoietin isoforms ANGPT¹ and ANGPT², similar to some angiopoietin-like proteins, exhibit multi-biological properties, including functional roles in different diseases, lipid metabolism,^16-21 inflammation,²² hematopoietic stem cell activity,^23-25 cancer cell invasion,^26,27 and diabetes and its complications.^28-49 To date, there is no reported study about the role of ANGPT¹, ANGPT², and angiopoietin-like proteins in PTDM. Kidney transplant is an invasive, expensive, compulsory procedure for all patients with end-stage kidney disease. To find a biomarker to identify kidney recipients prone to develop PTDM, we screened our cohorts for the above biomarkers, which are significant in diabetes and its complications.^28-49 In this study, we assessed the relationship between circulating angiopoietin proteins ANGPT¹ and ANGPT² and angiopoietin-like proteins ANGPTL³, ANGPTL⁴, ANGPTL⁶, ANGPTL⁷, and ANGPTL⁸ in kidney transplant recipients who developed PTDM (experimental group) versus recipients who did not develop diabetes after transplant (control group).

Materials and Methods
In this cross-sectional study, all patients were enrolled in the Dasman Diabetes Institute, Diabetes Education Department, and outpatient clinics of Hamad Al Essa Organ Transplant Centre between May 2015 and December 2016. All patients included in this study had no symptoms suggestive of diabetes pretransplant (especially those with idiopathic end-stage kidney disease) or its related complications, as shown with standard fasting and a 2-hour postprandial blood sugar test with a 2-hour oral glucose tolerance test. Moreover, blood sugar monitoring during the early posttransplant period was within reference range until after 6 months when they were diagnosed to have PTDM. A total of 309 renal transplant recipients aged 20 to 80 years participated in this study, with a mean age of 52.85 ± 11.4 years for recipients who developed PTDM versus 38.97 ± 13.1 years for kidney recipients who did not develop diabetes (control group). All biomarkers recruited in this study were assayed using enzyme-linked immunosorbent assay kits as previously reported.³¹ Signed informed consent forms were obtained from all participants. The research was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines and adhered to local and national regulatory requirements and laws. We excluded pediatric patients, pregnant women, patients with estimated glomerular filtration rate below 30 mL/min/m², and patients with mental retardation or unable to give consent.

Statistical analyses
We used SPSS software (version 26.0; IBM) for data analyses. The sample size was calculated to accept a marginal error of 6.6% (95% CI) in a normally distributed population. We compared variables and mean values with the paired-sample t test, independent sample t test, chi-square test, Fisher exact test, and analysis of variance, as appropriate. We expressed results as means ± SD. Differences were considered significant at P ≤ .⁰⁵. To observe effects of our sampled analytes, we studied risk factors such as BMI, sex, and age.

Results
The distribution of kidney transplant recipients who developed PTDM versus recipients who did not develop PTDM are shown in Table 1. Ages of subjects were distributed by age into <40 years, between 40 and 60 years, and >60 years. The mean of age was similar for donors in both groups. Our subjects were well distributed between Kuwaitis and non-Kuwaitis. The sex distribution (male vs female) was similar between the experimental and control groups (Table 1). Younger patients (<40 years old) were more prevalent in the control group, whereas patients >40 years were predominant in patients with new-onset diabetes mellitus after transplant (P < .⁰⁰¹). All patients received grafts from donors aged 30 to 40 years. The 2 groups were comparable in variability of original kidney disease, dialysis modality, donor source, and immunosuppression regimen (both induction and maintenance). Moreover, all subjects had no history of diabetes or symptoms suggestive of diabetes in addition to standard fasting and postprandial blood glucose levels prior to transplant.¹ There were no differences in ethnicity of our cohorts (55% Kuwaiti, 45% non-Kuwaiti). Moreover, no significant differences were noted in pretransplant comorbidities in both cohorts, especially hypertension, history tuberculosis treatment, ischemic heart disease, bone disease, anemia, hyperlipidemia, and human leukocyte antigen class I (A and B) and class II (DR) mismatches (P > .⁰⁵; Table 1). Patients with chronic hepatitis C virus infection were more prevalent in the PTDM group versus the control group (12 cases vs 2 cases; P = .⁰⁰⁶; Table 1). Obesity, as measured by mean basal body mass index, was significantly higher in the PTDM group versus the control group (28.07 ± 5.5 vs 26.11 ± 7, respectively; P = .⁰¹), whereas follow-up mean body mass index was comparable in both groups irrespective of age (P = .²¹; Table 1). We observed that ANGPT¹ was significantly associated with PTDM development (P < .⁰⁰¹), whereas ANGPT² did not show any significant association with PTDM development in our study participants (Figure 1). Isoforms ANGPTL³ and ANGPTL⁴ did not show statistical correlation with PTDM. However, as shown in Table 2, ANGPTL⁶ was associated with PTDM incidence (P = .⁰⁰¹), but ANGPTL⁷ and ANGPTL⁸ were negatively associated with PTDM incidence (P < .⁰⁰¹ and P = .⁰⁴, respectively). Figure 1 illustrates the association screening of angiopoietin protein and angiopoietin-like protein biomarkers for our study.

Discussion
Angiopoietins ANGPT¹ and ANGPT² are 2 major isoforms that standardize vascular homeostasis.^7-9 The β-cell-derived ANGPT¹ isoform regulates insulin secretion and glucose homeostasis by stabilizing the blood vessels in the islet and therefore may be a novel therapeutic target for diabetes management in the future.⁴⁹ In diabetes mellitus, chronic hyperglycemia leads to the accumulation of advanced glycation end products and mitochondrial overproduction of reactive oxygen species that cause the upregulation of ANGPT² mRNA, which has been reported to promote vascular permeability, destabilization, and sprouting, further inducing microvascular and macrovascular complications as has been noted by previously published in vitro and in vivo studies.^50-52 Therefore, ANGPT² is an independent predictor of adverse clinical outcomes in patients with diabetes. The ANGPT² isoform has been reported to be elevated in patients with diabetes and to be associated with indexes of endothelial damage/dysfunction.⁵³ In our study, we observed that the level of ANGPT¹ was negatively associated with PTDM, whereas the ANGPT² isoform did not show any statistical association with either of the study groups. Protective action of ANGPT¹, in contrast, has confirmed its crucial role in β-cell formation.⁴⁹ On the other hand, ANGPT², which has been shown to have a significant association with diabetes,^32,39,40 did not show any association with PTDM development. This is the first report about the association of ANGPT¹ with PTDM and lack of association of ANGPT² with PTDM. The difference of results could be attributed to the well-controlled diabetes among our patients with diabetes who did not show microvascular complications that required long-standing hyperglycemia to develop. Azadi and colleagues⁵⁴ have provided in vivo evidence for the relation of ANGPTL³ with renal dysfunction and hypertriglyceridemia in patients with diabetic nephropathy, which is in line with our experimental findings and suggests a potential role for this hepatokine in diabetic nephropathy pathogenesis. In addition, ANGPTL³ and ANGPTL⁴ were associated with diabetes and its complications.^33,36,46-48 In our study, ANGPTL³ and ANGPTL⁴ did not show any significant association with either of the PTDM cohorts. However, ANGPTL⁶ and ANGPTL⁸ showed a significant association with PTDM incidence, whereas ANGPTL⁷ showed protection against PTDM development. A high level of serum ANGPTL⁶, associated with a low incidence of diabetes in humans, is expressed and secreted in response to hyperglycemia to maintain glucose homeostasis.³⁵ Issa and colleagues⁵⁵ have highlighted a possible predictive role of ANGPTL⁸ in diabetic complications, particularly nephropathy. They recommended larger prognostic studies to validate the cause-effect relationship between ANGPTL⁸ and deteriorated kidney functions. Yang and colleagues⁵⁶ have provided evidence for a strong link between ANGPTL⁸ and albuminuria, signifying that ANGPTL⁸ may be a new biomarker for diabetic kidney disease in type 2 diabetes. Levels of ANGPTL⁸ were significantly higher in gestational diabetes (GDM) than in the healthy control group. Elevated levels of serum ANGPTL⁸ were correlated significantly with insulin resistance parameters, which represent the main component of GDM pathophysiology. As previously reported by Seyhanli and colleagues,⁵⁷ ANGPTL⁸ could be a new biomarker for diagnosis of GDM. Our results suggested that angiopoietin proteins and angiopoietin-like proteins function differently in type 2 diabetes and posttransplant diabetes. Our initial data suggested that ANGPT¹ and ANGPTL⁶, ANGPTL⁷, and ANGPTL⁸ could be good candidates for future therapeutic regimens among kidney transplant recipients with PTDM.

Conclusions
Posttransplant diabetes mellitus is an important complication of kidney transplant. In our kidney transplant recipients with PTDM, ANGPT² did not show any significant association with microvascular complications. The plasma levels of ANGPT¹ and ANGPTL⁷ were strongly protective against PTDM, suggesting a possible role for these 2 angiopoietin-like isoforms in glucose homeostasis and β-cell function. On the other hand, ANGPTL⁶ and ANGPTL⁸ showed a strong association with PTDM, whereas ANGPTL⁷ showed a strong protective effect against PTDM. With regard to this novel finding, we recommend more studies on the effects of ANGPT¹, ANGPTL⁶, ANGPTL⁷, and ANGPTL⁸ in PTDM. In addition, angiopoietin proteins and angiopoietin-like proteins could be novel therapeutic targets for PTDM in the future. Plasma levels of ANGPT¹, ANGPTL⁶, ANGPTL⁷, and ANGPTL⁸ may correlate with disease severity and chronicity in transplant patients who develop PTDM, and levels may also serve as potential biomarkers of PTDM disease severity with possible early intervention in the management strategy, especially with modifiable risk factors such as immunosuppression regimen, antidiabetic medications, and follow-up of complications.