Biosafety of purslane seed extracts

No animal mortality or acute signs of toxicity were observed during the 14-day experimental period, and both MO and MC extracts were safe up to 5 g/kg single dose. Therefore, a dose of 250 mg/kg was selected for repeated administration. No significant change in the organosomatic indices, or food and water intake (data not shown) was observed during or at the end of the dosing period. However, the MC-treated animals showed a moderate reduction of body weight relative to the sham animals starting from the 21st day of treatment (P = 0.0068–0.022), unlike the MO-treated animals that showed no significant difference along the experimental period. The repeated MO dosing produced significantly lower blood glucose levels (P = 0.0162), while the MC dosing exhibited a slight elevation of serum albumin (P = 0.0044). Both treatments exhibited a significant reduction of total cholesterol and triglycerides (P < 0.0001 / each), with a higher effect observed in MO-treated animals in the former (P < 0.0001), and a higher effect observed in MC-treated animals in the latter (P = 0.0003). The treated animals showed lower levels of ALT (P = 0.0001, 0.0033; respectively) and also AST despite the absence of statistical difference. As both extracts were investigated as renoprotective candidates, different parameters of renal functions were also assessed after repeated dosing. Results showed that both MO- and MC-treated groups had significantly lower levels of serum uric acid (P = 0.0001, 0.0024; respectively), and urinary creatinine (P < 0.0001 / each), while no significant difference was found in BUN, proteinuria, as well as creatinine clearance and eGFR rates. Meanwhile, mild elevation of serum creatinine (P = 0.0001, 0.0069), and urine volume (P = 0.0092, 0.0232) were observed in both MO- and MC-treated groups; respectively. Changes in body weight during the dosing period and the terminal values of organosomatic index and blood biochemistry parameters are listed in Supplementary Tables 2, 3, and 4; respectively.

MO and MC treatments ameliorate hyperglycemia-associated symptoms

All the classical signs of diabetes including elevated blood glucose levels accompanied by weight loss, polyphagia, polydipsia, and polyuria were significantly pronounced in the untreated diabetic animals in the DN control group. The sham animal group had steady FBG levels and a normal weight gain pattern consistent with food consumption rate along the study period, whereas untreated diabetic animals in the DN group showed significantly higher FBG levels and water intake/ food consumption rates with lower mean body weight in all time points compared with the sham animals (P < 0.0001 / each). The MO-treated diabetic rats demonstrated a rapid improvement regarding FBG levels, body weight, as well as food/water consumption after only 7 days of treatment, and maintained a steady pattern. Meanwhile, the improvement of DM-associated symptoms was observed in those receiving the MC extract and Losartan after 14 days of treatment. The untreated diabetic animals showed a 31.5% increment of terminal FBG levels, concomitant with 27.26% weight loss compared with the initial values (P < 0.0001 / each), while both MO and MC extracts produced 67.35 and 63.1% lower FBG levels; respectively compared with the untreated diabetic rats (P < 0.0001 / each) at the same time point. Compared with Losartan-treated animals, the terminal FBG levels showed no significant difference in MO-treated rats unlike those treated with MC that showed significantly higher FBG levels compared with both Losartan- and MO-treated animals (P < 0.0001, 0.0038; respectively), while the latter showed a significantly higher terminal body weight relative to both MC- and Losartan-treated animals (P < 0.0001 / each). Changes in body weight, blood glucose, and food/water intake in different animal groups are detailed in Table (1).

Table 1 Changes in body weight, blood glucose, and food/water intake in untreated and treated diabetic rats in different groupsMO and MC treatments ameliorate the altered renal functional and structural markers in diabetic animals

In line with progressive hyperglycemia, the early signs of renal injury were observed in the untreated diabetic rats between the 7th and 14th days after the STZ administration, in terms of higher serum creatinine (Fig. 1.A); uric acid (1.B), and BUN levels (1.C), in addition to significant polyuria (1.D), higher levels of albuminuria (1.E), proteinuria (1.F), and urinary creatinine (1.G) concomitant with lower creatinine clearance (1.H) and eGFR rates (1.I) compared with the sham animals (P < 0.0001/ each) in all time points. All treatments significantly improved the regression of renal functional and structural markers compared with the untreated animals. In detail, the MO-, MC-, and Losartan-treated animals showed lower levels of serum creatinine (P < 0.0001 / each), uric acid (P = 0.0022, 0.0052, < 0.0001; respectively), and BUN levels (P = 0.0265, 0.0338, 0.0187; respectively) compared with the untreated after 7 days of treatment with no significant difference observe between the three treated groups. Similarly, a significant reduction of the polyuria (P < 0.0001, 0.0402, < 0.0001; respectively), renal proteome including levels of albuminuria (P < 0.0001 / each) and proteinuria (P = 0.0054, 0.0003, 0.0193; respectively), and urinary creatinine (P < 0.0001 / each) was observed in the three treated groups after 7 days of treatment, which also extended to improved rates of creatinine clearance (P < 0.0001 / each) and eGFR (P = 0.0123, < 0.0001, 0.0385; respectively). As demonstrated in Fig. (1.D). The MO treatment showed a persistently higher effect on the polyuria in all time points recorded compared with MC and Losartan, while the MC exhibited significantly higher improvement of renal proteome and both creatinine clearance and eGFR rates relative to MO and Losartan (P < 0.01 / each) (1.E-I).

Fig. 1

Effect of MO and MC treatments on renal injury markers. Compared with the untreated control, all treatments significantly reduced the elevated levels of serum creatinine (P < 0.0001 / each) (A), uric acid (P = 0.0022 - <0.0001) (B), and BUN (P = 0.0187–0.0338) (C). The MO-, MC-, and Losartan-treated animals also showed improved urine parameters compared with the untreated animals, including a significant reduction of polyuria (P = 0.0402 - <0.0001) (D), renal proteome (P = 0.0193 - <0.0001) (E, F), urinary creatinine (P < 0.0001 / each) (G), and the rates of both creatinine clearance (P < 0.0001 / each) (H) and eGFR (P = 0.0385 - <0.0001) (I) within 7 days of treatment

MO and MC treatments halt the progression of diabetic nephropathy

The early markers of diabetic nephropathy were detected in the untreated diabetic animals (DN) after 14 days of STZ dosing, where the urinary levels of transferrin (Fig. 2.A); fibronectin (2.B); SDF-1 (2.D); IL-6 (2.F); TNF-α (2.H), and IL-18 (2.J) were found to be significantly higher than those in the sham and treated diabetic animals, concomitant with significant reduction of serum SDF-1 (2.C), and elevated IL-6 (2.E); TNF-α (2.G); IL-18 (2.I), and hs-CRP levels (2.K) at all time points (P < 0.0001 / each). The three treatments; particularly the MC extract, halted the progression of the DN as demonstrated by significant improvement of the dysregulated markers’ levels until they became near the baseline or undetectable levels by the end of the study. The onset of each treatment’s modulatory effect was observed at different time points, where the MO extract exhibited higher activity in reducing the elevated levels of urinary transferrin, urinary IL-18, and serum TNF-α by 3.31, 9.24, and 4.31% (P = 0.0036, 0.0003, 0.0454); respectively, while it upregulated serum SDF-1 levels by 5.27% (P = 0.0046) relative to MC extract after 7 days of treatment. Meanwhile, the latter exhibited superior regulatory effects on urinary SDF-1, serum/urinary IL-6, urinary TNF-α, and serum IL-18 at the same time point compared with MO extract, where treated animals had 4.3, 12.35, 45.63, 5.98, and 29.64% lower levels (P = 0.0091, < 0.0001, < 0.0001, 0.0109, 0.0001); respectively, while they had 9.7, 10.26, 42.16, and 43.91% lower levels of transferrin, serum TNF-α, urinary SDF-1, and hs-CRP levels (P < 0.0001, < 0.0001, 0.0008, < 0.0001, and 0.0001); respectively at the 14th day. Regarding fibronectin levels, no significant difference was observed in MC-treated animals relative to those treated with the MO extract before the 42th day, where the animals showed 84.83% (P < 0.0001) lower levels, while the MC treatment exhibited a higher inhibitory effect on urinary IL-18 levels at the 21st day of treatment, and the treated animals ended up with undetectable levels. Compared with Losartan, the MO-treated animals showed 2.29, 3.84, 19.9, and 17.75% (P = 0.0305, 0.0106, < 0.0001/each; respectively) lower transferrin, serum/urinary IL-6, urinary TNF-α; respectively at the 7th day of treatment, while the MC-treated animals showed 13.77, 18.47, 120.49, 25.24, and 24.04% (P = 0.0007, < 0.0001 / each; respectively) lower fibronectin, serum/urinary IL-6, urinary TNF-α, and serum IL-18; respectively at the same time point.

Fig. 2

Effect of MO and MC treatments on early markers of diabetic nephropathy. All treatments halted the progression of the DN reflected by the significant reduction of urinary transferrin (A), fibronectin (B), urinary SDF-1 (D), serum/urinary IL-6 (E, F), TNF-α (G, H), and IL-18 (I, J), in addition to serum hs-CRP (K), and the significant upregulation of serum SDF-1 levels (C) comparing with the untreated control (P < 0.0001 / each). The regulatory onset of each treatment was observed at different time points, where the MO extract showed the modulatory effect on urinary transferrin and IL-18 besides serum TNF-α and SDF-1 during the first week of treatment, while the MC extract showed higher activity on urinary SDF-1, serum/urinary IL-6, urinary TNF-α and serum IL-18 at the same time point. The ultimate values of investigated markers were almost normalized in MC-treated animals

MO and MC treatments maintain the structural integrity of diabetic kidneys

The absolute weights and hypertrophy index (Ki) of the untreated kidneys were significantly higher than those observed in the control (P < 0.0001 / each), while the MC extract normalized both the weight and Ki value in treated kidneys with no statistical difference observed relative to the control. Meanwhile, both MO and Losartan-treated groups showed moderately higher kidney weights (17.2, 22.44%; P = 0.0279, 0.0016; respectively) and Ki values (19.92, 23.62%; P = 0.0497, 0.0105; respectively) compared with the MC-treated despite the lower values relative to the untreated control (P < 0.0001 / each). The microscopic assessment of the renal sections obtained from the sham group (Fig. 3.A) revealed normal kidney architecture. Glomeruli (G) were intact with normal-sized basement membrane and Bowman’s (urinary) space, and both proximal (PCT) and distal convoluted tubules (DCT) were normal-appearing, with obvious vesicular round nuclei and granular cytoplasm. Similar findings were observed in sections obtained from control animals receiving only the MO (3.B) and MC extracts (3.C). In diabetic renal Sect.  (3.D), severe glomerular congestion associated with massive periglomerular cellular infiltrations was observed. The untreated glomeruli showed marked hypercellularity, with markedly collapsed tufts that resulted in urinary space widening. Both proximal and distal convoluted tubules showed marked degenerative changes including epithelial desquamation, necrosis, cellular swelling, intra-cytoplasmic vacuolization, and darkly stained pyknotic nuclei, with moderate hemorrhage observed in interstitial tissue. The MO treatment (3.E) alleviated the glomerular congestion and reduced the tubular degeneration compared with the untreated kidneys; however, moderate glomerular hypercellularity associated with widened urinary space (US) and mild periglomerular cellular infiltration were still detectable. Meanwhile, the MC-treated kidneys showed near to normal architecture with minute cellular infiltration, intact glomeruli with well-defined Bowman’s space, and intact convoluted tubules with mild degeneration and pyknotic nuclei (3.F). Notably, reformation of renal tubular epithelium was also observed. Similarly, the Losartan-treated kidneys showed normal architecture with few cellular infiltrations around glomeruli and mild degeneration of proximal (PCT) and distal (DCT) convoluted tubules and pyknotic nuclei (3.G). Renal injury scores based on microscopic assessments are listed in Table (2).

Fig. 3

Effect of MO and MC treatments on histopathological alterations in diabetic kidneys. Sections of normal kidneys (A) and control groups receiving only MO (B) and MC (C) extracts showed the normal-appearing architecture of the glomeruli (G); urinary space (US), Proximal (PCT), and distal (DCT) convoluted tubules. The untreated diabetic kidneys (D) showed severe glomerular and tubular alterations, including collapsed glomerular tuft (G) associated with wide urinary space (US), epithelial desquamation of proximal (PCT) and distal (DCT) convoluted tubules, cellular swelling, intra-cytoplasmic vacuolization (thin arrow), and pyknotic nuclei (arrowhead). MO treatment has markedly alleviated the periglomerular cellular infiltration in diabetic kidneys (E), therefore, the glomeruli were less congested and/or hypercellular, and merely mild degeneration was observed in proximal (PCT) and distal (DCT) convoluted tubules. The widened urinary space (US) (thin green arrows) was not normalized. Both MC- and Losartan-treated kidneys (F, G) showed normal structure of the glomerulus (G), and urinary space. (US) Proximal (PCT) and distal (DCT) convoluted tubules with few cellular infiltrations around glomeruli (thin arrow)

Table 2 Scoring of macro-/microscopic alterations in untreated and treated diabetic kidneys in different groupsMO and MC treatments restore the effectuality of the antioxidant system in diabetic kidneys

The accumulation of reactive oxygen associated with excessive lipid peroxidation was more pronounced in the untreated diabetic kidneys. Compared with sham animals, kidneys of untreated diabetic animals had 4.23 and 3.45 fold higher levels of total ROS and MDA; respectively (Fig. 4.A, B), concomitant with 51.7% reduction of GSH levels (4.C) and 50.6 and 51.57% lower activity of glutathione reductase (GR) and superoxide dismutase (SOD) (4.D, E) (P < 0.0001 / each). The three treatments exerted significant improvement of the dysregulated antioxidant markers with the MO treatment showing the highest ameliorative effect. The MO, MC, and Losartan-treated kidneys showed 48.41, 39.33, and 47.76% reduction of total ROS, and 56.62, 48.9, and 38.65% reduction of MDA levels; respectively relative to the untreated kidneys (P < 0.0001 / each). Also, the three treatments upregulated the tissue levels of GSH by 71.74, 59.05, and 65.4%; respectively, and promoted the activities of both GR and SOD by 79.25, 66.42%; 75.1, 54.22%, and 73.88, 45.24%; respectively, relative to the untreated diabetic kidneys (P < 0.0001 / each). Compared to Losartan-treated kidneys, no statistical difference was observed in MO-treated kidneys regarding the ROS, GSH levels, or GR activity, while showing 29.29% lower MDA (P < 0.0001) and 12.73% higher SOD activity (P = 0.0008). Meanwhile, the MC-treated kidneys showed 13.9% higher ROS (P = 0.006) and 16.71% lower MDA levels (P = 0.0004), but no statistical difference was observed regarding GSH levels, GR and SOD activities.

Fig. 4

Effect of MO and MC treatments on redox status in diabetic kidneys. Significant elevation of total ROS (A) and MDA production (B) was observed in the diabetic kidneys, concomitant with significant reduction of reduced GSH (C), and lower activity of the antioxidant enzymes showed glutathione reductase (GR) and superoxide dismutase (SOD) (D, E) when compared with normal kidneys (P < 0.0001 / each). The three treatments significantly suppressed the ROS accumulation and lipid peroxidation, while they significantly elevated the tissue levels of GSH, and promoted the activity of glutathione reductase (GR) and superoxide dismutase (SOD) in diabetic kidneys (P < 0.0001 / each). Notably, the highest improvement of distorted renal redox was observed in MO-treated kidneys

MO and MC treatments modulate the transcription of the oxido-inflammatory mediators in diabetic kidneys

Relative gene expression of different mediators of the oxido-inflammatory axis including NF-κB, IL-6, TNF-α, IL-10, Nrf2, Keap1, HO-1, and PPARγ was evaluated in diabetic kidneys using quantitative PCR (qPCR). Results revealed a significant upregulation of TNF-α and IL-6 expression by 3.25, 2.87 fold in the diabetic kidneys (P < 0.0001 / each), while the expression of IL-10 was significantly inhibited by 51.5% (P = 0.0034) compared with normal kidneys (Fig. 5.A). Associated with these events, the untreated kidneys showed 58.5, 60 and 66.7% reduction of Nrf2, HO-1, and PPARγ mRNA folds (P < 0.0001, < 0.0001, 0.0003); respectively, while the relative expression of Keap1 and NF-κB was upregulated by 1.625 and 7.738 fold (P < 0.0001 / each); respectively (5.B). The relative expression of TNF-α and IL-6 genes were significantly downregulated in MO-; MC-, and Losartan-treated kidneys by 76.92, 67.2%; 74.46, 58.18%, and 75.38, 51.22%; respectively (P < 0.0001 / each) compared with the untreated control. Meanwhile, all treated kidneys showed significant upregulation of IL-10 expression by 53.6, 80.2, and 52.34% (P = 0.0002, < 0.0001, 0.0004); respectively. Concomitantly, the expression of Nrf2, HO-1, and PPARγ were significantly upregulated by 79.37, 65.32, 65.8%; 71.46, 50.82, 80.2%, and 71.35, 41.18, 67.14%in MO-, MC-, and Losartan- treated kidneys (P < 0.0001 / each); respectively, whereas the expression of Keap1 and NF-κB was significantly inhibited by 60, 68.44% (P < 0.0001 / each); 44.61, 78.91% (P < 0.0001 / each), and 20, 67.7% (P = 0.0004, < 0.0001); respectively. No significant difference in TNF-α expression level in Losartan-treated kidneys compared with both MO- and MC-treated kidneys while showing 21.2, 63.87%; 38.32, 61.31%, and 50, 30.77% higher COX2, PGE2, and Keap1 expression levels (P < 0.0001 / all), and 35.5, 10.78%; 41.04, 16.39% lower Nrf2 (P < 0.0001, 0.0189) and HO-1 (P < 0.0001, 0.0357); respectively. Unlike the MC-treated kidneys, the MO-treated showed 32.85% lower IL-6 expression (P = 0.0037) compared with those treated with Losartan, while no significant difference was detected regarding the relative expression of IL-10, NF-κB, and PPAR-γ. The MC-treated kidneys; however, showed 58.42 and 39.76% higher IL-10 and PPAR-γ expression levels, 33.17% lower NF-κB (P < 0.0001 / all) compared with the Losartan-treated, with no difference observed in the MO-treated.

Fig. 5

Effect of MO and MC treatments on the gene expression of oxido-inflammatory mediators in diabetic kidneys. Compared with normal kidneys, the diabetic kidneys showed 3.25, 2.87, 1.625, and 7.738 fold higher expression of TNF-α, IL-6, Keap1, and NF-κB genes, while they had 51.5, 58.5, 60 and 66.7% lower IL-10, Nrf2, HO-1, and PPARγ expression (P = 0.0034 - <0.0001). Both MO and MC extracts as well as Losartan treatments significantly downregulated the expression of TNF-α, IL-6, Keap1, and NF-κB genes, while upregulating the transcription of IL-10, Nrf2, HO-1, and PPARγ (P = 0.0004 - <0.0001). The highest expression levels of the Nrf2/HO-1 axis were observed in MO-treated kidneys, while the highest expression of PPARγ was observed in those treated with the MC extract

To elucidate the signaling cascades potentially implicated in the renoprotective effect of the two purslane extracts, the correlation pattern between hyperglycemia and associated oxidative stress with the expression levels of different oxido-inflammatory mediators in diabetic kidneys was analyzed. Results revealed strong positive correlations between both FBG and ROS levels with the mRNA folds of TNF-α, IL-6, Keap1, and NF-κB, while they were negatively correlated with those of Nrf2, HO-1, PPARγ, and IL-10 (Supplementary Table 5). In contrast, the serum level of SDF-1 showed an inverted correlation pattern with the same parameters; i.e. positively correlated with Nrf2, HO-1, PPARγ, and IL-10, while negatively correlated with TNF-α, IL-6, Keap1, and NF-κB mRNA folds. On the other hand, the expression level of Nrf2 was found to be positively correlated with that of HO-1, PPARγ, and to a lower extent with IL-10, while negatively correlated with TNF-α, IL-6, and Keap 1. Also, the expression profile of PPARγ showed a strong positive correlation with that of IL-10 while being negatively correlated with both TNF-α and IL-6, unlike the NF-κB which displayed an inverted correlation pattern with the same parameters.