非運鐵蛋白結合鐵及麩胱甘肽過氧化物酶4與犬貓慢性腎病之關係

Abstract

當系統性鐵過量或鐵代謝異常造成鐵的累積超出腎臟所能負荷，非運鐵蛋白結合鐵(non-transferrin-bound iron, NTBI)將沉積於腎小管細胞中。NTBI，也稱為催化鐵(catalytic iron)，能藉由芬頓反應(Fenton reaction)及類芬頓反應(Fenton-like reaction)在腎小管中產生活性氧化物質(Reactive oxygen species, ROS)，進而造成氧化傷害與細胞死亡。此種由鐵媒介的特殊細胞死亡方式被命名為鐵依賴性細胞死亡(ferroptosis)。麩胱甘肽過氧化物酶4 (Glutathione peroxidase 4, GPX4)為已知的唯一一種能減少脂質過氧化物的酶，因此GPX4也是ferroptosis的重要調節因子之一。血中NTBI已被證實在人類晚期腎病與腎病動物模型中均會上升，而GPX4的抑制則參與了缺血性腎損傷與腎臟纖維化的機制。然而，在小動物臨床醫學領域，GPX4、NTBI與腎臟疾病的關聯性仍未被研究過。本研究目的在於探討犬貓血中NTBI與尿中GPX4之關連性，及兩者與腎臟病程、臨床數值之相關，及是否能作為腎病進展的預測因子。 本研究共收集15隻健康貓、64隻慢性腎病貓、6隻慢性腎病合併急性腎損傷的貓，21隻健康犬、53隻慢性腎病犬，及11隻慢性腎病合併急性腎損傷犬隻。其中每個病患均完整記錄病史、臨床病理學檢查結果，並進行血清NTBI與尿中GPX4的檢測。血清NTBI使用感應耦合電漿質譜儀(inductively coupled plasma mass spectrometry, ICP-MS)進行偵測，而尿中GPX4則使用商品化酵素免疫分析套組(Enzyme-linked Immuno-sorbent Assay, ELISA)進行測量。尿中GPX4以尿肌酸酐(urine creatinine)濃度校正後，定義為urine-GPX4-to-creatinine ratio (UGCR); 血清NTBI以血容比(hematocrit)校正後，定義為serum NTBI-to-hematocrit ratio (SNHR)。 實驗結果顯示，在貓，UGCR在腎病功能惡化時會顯著提升(中位數[四分位距]: 控制組為0.072[0.057-0.101] ×10-4; 早期慢性腎病組為0.134[0.070-0.260] ×10-4; 晚期慢性腎病組為0.271[0.197-0.457] ×10-4, P <.001)。此外，當慢性腎病貓合併急性腎損傷(0.598[0.462-0.834] ×10-4)時，UGCR也會顯著較慢性腎病組高(0.199[0.118-0.328] ×10-4, P <.001)。在慢性腎病次分組中，UGCR在蛋白尿組(0.436[0.272-0.546] ×10-4)顯著較無蛋白尿組高(0.182[0.110-0.266] ×10-4, P <.001)，在高血壓組(0.407[0.262-0.512] ×10-4)也較無高血壓組提升(0.154[0.102-0.279] ×10-4 , P <.001)，並且，UGCR在貧血組(0.440[0.286-0.512] ×10-4)較無貧血組高(0.193[0.112-0.277 ×10-4, P =.001])，此外，在使用鐵補充劑組(0.435[0.214-0.616] ×10-4)也相較於無使用鐵補充劑組高(0.194[0.110-0.275] ×10-4 , P =.003)。UGCR (odds ratio [OR]: 1.419, P=.006)與SNHR (OR: 1.171, P= .017)均為慢性腎病在90天內進展的風險因子。在Cox比例風險模型中，SNHR >0.230(hazard ratio[HR], 16.37; 95% CI, 1.05-254.98; P = .046)或是UGCR>0.218(HR, 4.61; 95% CI, 1.07-19.87; P = .004)的慢性腎病貓，較有可能在90天內發生腎病的進展。在分類與迴歸樹(classification and regression tree)與階層式分群法分析(hierarchical clustering analysis)中，SNHR相較於其他臨床因子，對於90天內腎病進展為最有優勢的因子。 在犬隻，UGCR在晚期腎病組(0.703[0.214-1.614] ×10-4)也較控制組(0.063[0.040-0.090] ×10-4)或早期腎病組(0.112[0.057-0.346] ×10-4, P <.001)為高。在慢性腎病合併急性腎損傷組，UGCR (0.701[0.310-1.444] ×10-4)相較控制組(0.063[0.040-0.090] ×10-4, P =.001)顯著提升，而SNHR (0.219[0.210-0.236] ppb ×10-2)在該組則顯著高於控制組(0.139[0.118-0.152] ppb ×10-2)與慢性腎病組(0.164[0.135-0.190] ppb ×10-2, P=.002)。在慢性腎病次分組分析中，UGCR在蛋白尿組(0.531[0.171-1.179] ×10-4)顯著高於無蛋白尿組(0.082[0.042-0.105] ×10-4, P =.001)，且在高血壓組(0.531[0.164-1.062] ×10-4)也顯著高於無高血壓組(0.120[0.059-0.444] ×10-4, P =.011)，另外，SHNR在蛋白尿組(0.186[0.157-0.217] ppb ×10-2)顯著高於無蛋白尿組(0.142 [0.091-0.167] ppb ×10-2, P =.02)。另一方面，UCGR (P =.008)與SNHR (P =.001)與血中尿素氮均呈現線性相關。然而，UCGR (P= .48)或SNHR (P= .55)對於慢性腎病的進展均非顯著的風險因子。 總結來說，UGCR在犬貓均與腎臟功能的惡化相關，在晚期腎病犬貓均可見UGCR的上升。此外，在併發急性腎損傷的慢性腎病貓隻中，可見UGCR顯著比慢性腎病貓更高。UGCR與SNHR之間的機制可能具有種別差異，在貓隻兩者具有線性相關，然而在狗沒有顯著相關。而SNHR在慢性腎病貓隻中可用於預測腎病進展。

As systemic iron overload or iron metabolism abnormality has overpassed the capacity of iron storage in kidney, non-transferrin-bound iron (NTBI) accumulates in tubular cells. NTBI, also known as catalytic iron, generates reactive oxygen species (ROS) via Fenton reaction and Fenton-like reaction, which induces oxidative damage in renal molecules, and ultimately cell death. This unique modality of cell death is termed as ferroptosis. Glutathione peroxidase 4 (GPX4), the only enzyme capable of reducing lipid peroxidation products in cells, is reported to be a main regulator of ferroptosis. NTBI has been identified in people with end-stage renal disease and animal models with renal diseases. GPX4 inactivation is involved in post-ischemic kidney injury and kidney fibrosis. However, the role of NTBI and GPX4 in cats and dogs with chronic kidney disease (CKD) has not been previously reported. The aim of this study was to discover whether serum NTBI and Urinary GPX4 were associated with each other in cats and dogs with CKD. Besides, whether serum NTBI and Urinary GPX4 will be potential prognostic factors of CKD in small animal medicine would also be evaluated. Fifteen healthy cats, 64 cats with CKD, 6 cats with A-on-CKD, 21 healthy dogs, 53 dogs with CKD, and 11 dogs with A-on-CKD were enrolled. Complete patient history, clinicopathologic tests, and results of serum NTBI and urinary GPX4 were recorded in each case. Urinary GPX4 was measured by commercial ELISA kits for cats and dogs. Serum NTBI is measured by inductively coupled plasma mass spectrometry (ICP-MS). After being adjusted by urine creatinine, urine-GPX4-to-creatinine ratio (UGCR) was determined. Besides, serum NTBI-to-hematocrit ratio (SNHR) was defined as serum NTBI divided by hematocrit. In cats, UGCR increased significantly along with the decline of renal function. (median [interquartile range]: 0.072[0.057-0.101] ×10-4 in control group; 0.134[0.070-0.260] ×10-4 in early-stage group; 0.271[0.197-0.457] ×10-4 in late-stage group, P <.001) Additionally, UGCR in A-on-C group was significantly higher than in the CKD group. (0.598[0.462-0.834] ×10-4 in A-on-C group; 0.199[0.118-0.328] ×10-4 in CKD group, P <.001) In the subgroups of CKD, UGCR levels were significantly elevated in cats with proteinuria (0.436[0.272-0.546] ×10-4) compared to non-proteinuria (0.182[0.110-0.266] ×10-4, P <.001), in cats with hypertension (0.407[0.262-0.512] ×10-4) compared to non-hypertension (0.154[0.102-0.279] ×10-4 , P <.001), in cats with anemia (0.440[0.286-0.512] ×10-4) compared to non-anemia (0.193[0.112-0.277 ×10-4, P =.001]), and in cats taking iron supplements (0.435[0.214-0.616] ×10-4) compared to no iron supplements (0.194[0.110-0.275] ×10-4 , P =.003). UGCR had a positive linear association with SNHR. (P =.001) Both UGCR (odds ratio [OR]: 1.419, P=.006) and SNHR (OR: 1.171, P= .017) were risk factors of 90-day progression of CKD. In Cox proportional hazard model, cats with either SNHR >0.230(HR, 16.37; 95% CI, 1.05-254.98; P = .046), or UGCR>0.218 (HR, 4.61; 95% CI, 1.07-19.87; P = .004) had higher probabilities of progression within 90 days. In classification and regression tree (CART) algorithm and hierarchical clustering analysis, SNHR was the optimal parameter associated with 90-day progression among other items. In dogs, UGCR was also significantly increased in late-stage CKD (0.703[0.214-1.614] ×10-4) compared to the control group (0.063[0.040-0.090] ×10-4) and early-stage group (0.112[0.057-0.346] ×10-4, P <.001). In the A-on-C group, UGCR (0.701[0.310-1.444] ×10-4) was elevated compared control group (0.063[0.040-0.090] ×10-4, P =.001), while SNHR (0.219[0.210-0.236] ppb ×10-2) was significantly higher than CKD group (0.164[0.135-0.190] ppb ×10-2) and control group (0.139[0.118-0.152] ppb ×10-2, P=.002). Among the subgroups, UGCR was increased in dogs with proteinuria (0.531[0.171-1.179] ×10-4) compared to non-proteinuria (0.082[0.042-0.105] ×10-4, P =.001), and dogs with hypertension (0.531[0.164-1.062] ×10-4) compared to non-hypertension (0.120[0.059-0.444] ×10-4, P =.011). Besides, SNHR was elevated in dogs with proteinuria (0.186[0.157-0.217] ppb ×10-2) compared to non-proteinuria (0.142 [0.091-0.167] ppb ×10-2, P =.02). UGCR (P =.008) and SNHR (P =.001) were both positively correlated with blood urea nitrogen (BUN). However, neither UGCR (P= .48) nor SNHR (P= .55) was significant risk factors for 90-day progression of CKD in dogs. In conclusion, feline and canine UGCR were strongly associated with renal function. There was a significant elevation of UGCR in late-stage CKD patients. Besides, in cats with acute on CKD, UCGR increased even higher than the CKD cats. There were possible species differences in the mechanism of UGCR and SNHR. In cats, a significant linear correlation was discovered between UGCR and SNHR, while no correlation was found in dogs. In addition, SNHR was a valuable predictor for CKD of 90-day progression in cats.