1. health and fitness
2. disease

Articles in PresS. Am J Physiol Renal Physiol (November 5, 2014). doi:10.1152/ajprenal.00555.2014
1
Glycosuria-mediated urinary uric acid excretion in patients with
2
uncomplicated type 1 diabetes mellitus
3
4
Yuliya Lytvyn, HBSc1,2, Marko Škrtić, MD PhD1, Gary K. Yang, PhD1, Paul M. Yip, PhD3,
5
Bruce A. Perkins MD4, David Z. I. Cherney, MD, PhD1
6
7
1
8
Toronto, Ontario, Canada
9
2
10
3
11
Toronto
12
13
14
15
16
4
17
18
19
20
21
22
23
24
25
26
27
28
29
Word Count: 3280 for main text and 247 for abstract; 1 table and 4 figures
Department of Medicine, Division of Nephrology, Toronto General Hospital, University of Toronto,
Department of Pharmacology and Toxicology, University of Toronto, Canada
University Health Network, Department of Laboratory Medicine and Pathobiology, University of
Department of Medicine, Division of Endocrinology, Mount Sinai Hospital, University of Toronto,
Toronto, Ontario, Canada
Running Title: Glycosuria causes uricosuria in type 1 diabetes
Please address correspondence to:
David Cherney, MD CM, PhD, FRCP(C)
Toronto General Hospital
585 University Ave, 8N-845
Toronto, Ontario,
M5G 2N2
Phone: 416.340.4151
Fax: 416.340.4999
Email: [email protected]
1
Copyright © 2014 by the American Physiological Society.
30
ABSTRACT:
31
Objective: Plasma uric acid (PUA) is associated with metabolic, cardiovascular and renal
32
abnormalities in patients with type 2 diabetes, but is less well understood in type 1 diabetes
33
(T1D). Our aim was to compare PUA levels and fractional UA excretion (FEUA) in patients with
34
T1D vs. healthy controls (HC) during euglycemia and hyperglycemia.
35
Methods: PUA, FEUA, blood pressure (BP), glomerular filtration rate (GFR-inulin) and effective
36
renal plasma flow (ERPF – paraaminohippurate) were evaluated in patients with T1D (n=66)
37
during clamped euglycemia (glucose 4-6 mmol/L) and hyperglycemia (9-11 mmol/L), and in HC
38
(n=41) during euglycemia. To separate the effects of hyperglycemia vs. increased glycosuria,
39
parameters were evaluated during clamped euglycemia in a subset of T1D patients before and
40
after SGLT2 inhibition for 8 weeks.
41
Results: PUA was lower in T1D vs. HC (228±62 µmol/L vs. 305±75 µmol/L, p<0.0001). In
42
T1D, hyperglycemia further decreased PUA (228±62 µmol/L to 199±65 µmol/L, p<0.0001),
43
which was accompanied by an increase in FEUA (7.3±3.8 to 11.6±6.7, p<0.0001). In T1D, PUA
44
levels correlated positively with SBP (p=0.029) and negatively with ERPF (p=0.031) and GFR
45
(p=0.028). After induction of glycosuria with SGLT2 inhibition while maintaining clamped
46
euglycemia, PUA decreased (p<0.0001) and FEUA increased (p<0.0001).
47
Conclusions: PUA is lower in T1D vs. HC, and positively correlates with SBP and negatively
48
with GFR and ERPF in T1D. Glycosuria rather than hyperglycemia increases uricosuria in T1D.
49
Future studies examining the effect of UA lowering therapies should account for the impact of
50
ambient glycemia, which causes an important uricosuric effect.
51
52
2
53
INTRODUCTION:
54
Humans have higher uric acid (UA) levels in comparison to other mammals due to
55
mutational silencing of the enzyme uricase, which results in UA remaining the end product of
56
purine metabolism (29). Additionally, about 90% of filtered UA is reabsorbed by the S1 segment
57
of the proximal convoluted tubule, in a process regulated by intracellular anion transporters on
58
the basolateral membrane, such as urate transporter 1 (URAT1) and the more recently discovered
59
glucose transporter 9 (GLUT9) (29). The lack of uricase, combined with the high reabsorptive
60
capacity in the kidney, predisposes humans to the development of hyperuricemia.
61
UA has recently emerged as an inflammatory factor that increases oxidative stress and
62
promotes activation of the renin angiotensin aldosterone system (RAAS) (29). From a clinical
63
perspective, higher UA levels are associated with metabolic abnormalities (insulin resistance,
64
hyperglycemia), cardiovascular disease (hypertension, endothelial dysfunction, arterial stiffness,
65
cardiac diastolic dysfunction) and kidney injury (28, 29) and thus could be involved in the onset
66
and progression of diabetic nephropathy, a common microvascular complication of diabetes
67
mellitus (DM). Plasma UA (PUA) levels could therefore serve as a biomarker and an effective
68
therapeutic target to supplement current clinical targets such as hemoglobin A1c (HbA1c),
69
cholesterol and blood pressure.
70
Evidence from rodent models suggests an association between high UA levels and
71
markers of high intraglomerular pressure such as hyperfiltration, and with subsequent increases
72
in proteinuria, glomerular sclerosis and tubulointerstitial fibrosis, leading to chronic kidney
73
disease (28). More recently, in an animal model of type 1 DM (T1D), UA lowering reduced
74
proteinuria, preserved GFR and suppressed renal expression of inflammatory interleukins (37).
75
In patients with T1D, UA is associated with impaired renal function, even when UA levels are in
76
the normal range (28, 35). For example, in 355 T1D participants from the second Joslin Study on
3
77
the Natural History of Microalbuminuria, baseline UA (within the normal range) showed a
78
significant association with early GFR loss of more than 3.3% per year over a 6-year follow-up
79
period (15). UA also increases the risk of developing proteinuria in T1D patients (23). For
80
example, in 652 normoalbuminuric type 1 DM patients recruited into the Coronary Artery
81
Calcification in Type 1 Diabetes Study (CACTI), each 60 μmol/L increment in UA from baseline
82
increased the risk of micro- or macro-albuminuria by 80% after a 6-year follow-up period (23).
83
Though observational associations between higher UA levels and renal outcomes show
84
consistency among independent cohorts (29), UA levels are not clearly defined in the T1D
85
populations.
86
Accordingly, the first goal of this study was to compare PUA levels in healthy control
87
patients (HC) with patients with T1D. It was hypothesized that even within the normal range,
88
PUA levels would be higher in the T1D cohort and that higher PUA levels will be associated
89
with deleterious hemodynamic profiles such as higher blood pressure and changes in renal
90
hemodynamic function. The second goal was to examine the relationship between clamped
91
hyperglycemia, hemodynamic parameters and PUA levels to determine if this acute
92
physiological stimulus, which promotes deleterious hemodynamic effects such as increased
93
blood pressure, influences PUA levels.
94
95
RESEARCH DESIGN AND METHODS:
96
Subject Inclusion Criteria and Study Preparation:
97
Forty-one HC and 66 T1D patients underwent detailed physiological examinations. In
98
brief, inclusion criteria were: 18-40 years of age, blood pressure <140/90, normoalbuminuria on
99
a 24-hour urine collection, diabetes duration >1 year, no history of renal or cardiovascular
4
100
complications and no intake of concomitant medications that would alter blood pressure or
101
cardiovascular outcomes. Study visits were performed after a controlled diet for 7 days
102
consisting of 150 mmol/day sodium and 1.5 g/kg/day protein. The sodium-replete diet was
103
used to avoid circulating volume contraction, RAAS activation and between-subject
104
heterogeneity. Pre-study protein intake was modest to avoid the hyperfiltration effect of high
105
protein diets (24). All the studies were approved by the University Health Network Research
106
Ethics Board and all subjects gave written informed consent.
107
108
Experimental Procedures:
109
Patients with T1D were studied on 2 consecutive days during euglycemia and
110
hyperglycemia. Euglycemic (4-6 mmol/L) and hyperglycemic (9-11 mmol/L) conditions in T1D
111
were maintained using a modified glucose clamp technique as previously described (8). Blood
112
glucose levels were stable for at least 2 hours prior to the measurement of the study end points
113
and were maintained 3 to 5 hours for the rest of the study day. HC were studied during
114
normoglycemic conditions at the Renal Physiology Laboratory at the Toronto General Hospital.
115
Glomerular filtration rate (GFR) and effective renal plasma flow (ERPF) were estimated using
116
inulin and paraaminohippurate (PAH) steady state infusion clearance techniques (5),
117
respectively, using previously described methods (9). The results of the 2 clearance periods were
118
averaged. Brachial artery blood pressure measurements were obtained at 30-minute intervals
119
throughout the study days (Critikon, Tampa, Florida, USA).
120
In a post-hoc analysis undertaken to understand the relative effects of hyperglycemia vs.
121
increased glycosuria, the effect of sodium glucose co-transporter 2 (SGLT2) inhibition on PUA
122
and urinary UA was examined using frozen, archived samples. The aim of this analysis was to
5
123
induce glycosuria while maintaining euglycemia to determine whether effects on PUA were due
124
to increased urinary UA excretion. For this analysis, we analysed urine and plasma samples
125
(n=40) obtained during baseline clamped euglycemic conditions and at follow-up after treatment
126
with empagliflozin 25mg QD for 8 weeks in the Adjunctive-To-Insulin and Renal MechAnistic
127
pilot trial of empagliflozin in T1D (ATIRMA trial, ClinicalTrials.gov NCT01392560). The
128
primary and secondary outcomes from this trial have been published (9).
129
130
Sample Collection and Analytical Methods:
131
After clamped euglycemia was achieved for at least 2 hours, blood was collected for
132
measurements of inulin, PAH, sodium, PUA and RAAS mediators (angiotensinogen, plasma
133
renin activity [PRA], aldosterone and angiotensin II) and urine samples were collected for UA,
134
sodium, glucose and creatinine measurements.
135
The blood samples were immediately centrifuged at 3000 rpm at 4ºC for 10 minutes.
136
Plasma was extracted, placed on ice and stored at -70ºC. Inulin and PAH were measured in
137
serum by colorimetric assays using anthrone and N- (1-naphthyl) ethylene-diamine respectively
138
(16). All hemodynamic measurements were adjusted for body surface area. Filtration fraction
139
(FF) represented the ratio of GFR to ERPF. Renal blood flow (RBF) was derived as ERPF / (1-
140
hematocrit). Renal vascular resistance (RVR) was derived by dividing mean arterial pressure
141
(MAP) by the RBF.
142
Plasma and urine samples were measured for UA, sodium, creatinine, glucose and urea
143
on the Architect c8000 Clinical Chemistry System using the manufacturer’s reagents (Abbott
144
Diagnostics, Abbott Park, Illinois, USA). In addition, UA excretion was expressed as fractional
145
excretion (FEUA), derived using (UUA×PCr)/(UCr×PUA)×100 where UUA, PCr, UCr and PUA are
6
146
urinary UA, plasma creatinine, urinary creatinine and plasma UA concentrations respectively.
147
Similarly, sodium excretion was expressed as fractional excretion (FENa), derived using
148
(UNa×PCr)/(UCr×PNa)×100 where UNa and PNa are urinary sodium and plasma sodium
149
concentrations, respectively.
150
Aldosterone was measured using a Coat-A-Count radioimmunoassay system. PRA was
151
measured using a radioimmunoassay kit (Diasorin, Stillwater, Minnesota, USA). HbA1c was
152
measured by high-performance liquid chromatography with the Variant II system (Bio-Rad
153
Laboratories, Hercules, California, USA).
154
155
Statistical analysis:
156
Data are presented as mean ± standard deviation (SD). To assess for between-
157
group differences, analysis of variance with post-hoc Tukey’s test was used. To compare within-
158
group differences (responses to hyperglycemia or SGLT2 inhibition) a paired student’s t-test was
159
used. Linear regression analysis was used to determine correlations between responses and PUA
160
levels. Statistical significance was defined as p<0.05. All statistical analyses were performed
161
using SAS v9.1.3 and GraphPad Prism software (version 6.0).
162
163
RESULTS:
164
Baseline characteristics
165
166
Baseline parameters were similar between HC and T1D patients (Table 1). Participants
were young, normotensive, normoalbuminuric and the two groups were similar in age and BMI.
167
During euglycemia, heart rate was significantly higher, but still within the normal range,
168
in the T1D versus HC and no significant differences in SBP or DBP were observed. During
7
169
hyperglycemic conditions, SBP significantly increased and HR decreased compared to
170
euglycemia in the T1D group. As expected, T1D participants had significantly lower levels of
171
circulating RAAS mediators compared to HC (aldosterone, PRA and angiotensin II) (41). During
172
hyperglycemia, aldosterone and PRA levels further decreased.
173
As expected, T1D subjects exhibited higher GFR, ERPF, RBF and lower RVR compared
174
to HC (p<0.0001 for all comparisons). Out of the 66 T1D patients, 29 exhibited normofiltration
175
(44%) and 37 hyperfiltration (56%), where hyperfiltration was defined as GFR 135
176
mL/min/1.73m2. In response to clamped hyperglycemia, GFR tended to increase in T1D (147±40
177
to 159±39 mL/min/1.73m2, p=0.064) and RVR decreased (0.069±0.021 to 0.055±0.016
178
mmHg/L/min, p<0.0001). No significant changes to ERPF, FF or RBF were observed in
179
response to hyperglycemic conditions.
180
181
Sodium, Glucose and UA Handling at Baseline
182
During clamped euglycemic conditions, PUA levels were lower in the T1D group vs. HC
183
(228±62 vs 305±75 µmol/L, p<0.0001) (Table 1, Figure 1). PUA negatively correlated with
184
FEUA in T1D patients (r=-0.60, p<0.0001). Uglucose excretion levels were also greater in T1D vs.
185
HC during clamped euglycemia, but there was no significant difference in urine uric
186
acid/creatinine ratio, FENa or FEUA between HC and T1D.
187
Compared to levels during clamped euglycemia, PUA decreased further in response to
188
clamped hyperglycemia (228±62 µmol/L to 199±65 µmol/L, p<0.0001) (Table 1, Figure 1). The
189
decline in PUA levels in T1D patients during hyperglycemia was accompanied by significant
190
increases in urine UA levels (257±121 to 339±161 umol/L, p=0.0007) and FEUA (7.3±3.8 to
191
11.6±6.7, p<0.0001). PUA was negatively correlated with FEUA during clamped hyperglycemia
8
192
(r=-0.50, p<0.0001). The increase in UA excretion during hyperglycemia was accompanied by
193
significant increases in Uglucose (1.4±3.2 to 9.8±10.4 mmol/L, p<0.0001) and FENa (0.87±0.56 to
194
1.63±0.89, p<0.001).
195
196
UA Correlations with Hemodynamic Parameters
197
PUA levels were positively correlated with SBP in T1D (r=0.27, p=0.029) under
198
euglycemic conditions, but not during hyperglycemia (Figure 4). During euglycemia, PUA levels
199
negatively correlated with ERPF (r=-0.27, p=0.031), and FEUA positively correlated with ERPF
200
(r=0.30, p=0.017) in T1D patients. During hyperglycemia, PUA negatively correlated with GFR
201
in T1D (r=-0.27, p=0.028).
202
203
Sodium, Glucose and Uric Acid Handling Upon Empagliflozin SGLT2 Inhibition
204
SGLT2 inhibitors are a new class of agents for the treatment of T2D that block proximal
205
renal tubular glucose reabsorption, leading to increased glucose excretion. Therapeutically, this
206
translates into important plasma glucose lowering effects (6). Trials with SGLT2 inhibitors in
207
patients with T2D have reported consistent and clinically relevant decreases in PUA levels (14,
208
40); however the mechanisms responsible were never clearly elucidated. Accordingly, to better
209
understand whether PUA lowering with hyperglycemia is due to systemic effects leading to
210
decreased production or renal effects causing increased uricosuria, plasma and urine UA levels
211
were measured before and after SGTL2 inhibition while maintaining clamped euglycemia in 40
212
T1D patients.
213
During clamped euglycemic conditions, after empagliflozin treatment, the anticipated
214
increase in Uglucose/creatinine (1.3±3.2 to 42.9±17.8, p<0.0001) was accompanied by a decline in
9
215
PUA (225±65 to 191±62 mmol/L, p<0.0001) and increases in UUA/creatinine (290±110 to 327±103
216
mmol/mmol, p=0.0075) and FEUA (8.2±3.6 to 11.1±5.1, p<0.0001) (Figure 2).
217
218
DISCUSSION:
219
Observational associations between higher UA levels and metabolic abnormalities,
220
cardiovascular disease and kidney dysfunction show consistency among independent healthy and
221
disease state cohorts, in both animals and humans (29). The potential renal protective effects of
222
UA lowering in T1D patients are being studied as part of the NIH funded Protecting Early Renal
223
Function Loss or “PERL” study (NCT02017171) (30), highlighting the promising future role for
224
UA-based therapies in T1D. However UA levels during euglycemia compared to hyperglycemia
225
have not been clearly defined in otherwise healthy T1D patients. Our first goal was to compare
226
PUA levels in HC with levels in patients with T1D. Our second goal was to determine if acute
227
clamped hyperglycemia, which promotes deleterious hemodynamic effects such as increased
228
blood pressure, influences PUA levels.
229
Due to the strong association between PUA levels and cardiovascular and renal
230
abnormalities, especially in the context of diabetes, it was initially hypothesized that T1D
231
patients would have higher PUA levels compared to HC. Our first major observation, however,
232
was that T1D patients had lower PUA levels under euglycemic conditions compared to HC, in
233
conjunction with increased urinary glucose that did not correlate with the degree of UA
234
excretion. Hyperglycemia in T1D patients was associated with a significant increase in urinary
235
sodium, glucose and UA excretion and thus a further PUA decrease, highlighting an important
236
physiological link between renal handling of UA, glucose and sodium (6). Furthermore, the
237
negative correlation between PUA and FEUA during euglycemia and hyperglycemia suggests that
10
238
PUA decreased as a result of increased renal excretion. The lack of elevated UA excretion in
239
T1D compared to HC under euglycemic conditions may suggest that T1D patients produce less
240
UA in plasma or consume less UA-containing products, although the similar protein intake based
241
on urine urea excretion in these groups suggests that differences in intake of UA-containing
242
foods were not relevant to our findings.
243
Our observations support several studies showing an increase in UA excretion in
244
response to intravenous D-glucose infusion (4, 36). More recently, an association was found
245
between lower PUA and poor glycemic control (11, 18, 22). Previous studies have shown that
246
insulin levels are positively correlated with PUA and insulin administration decreases UA
247
excretion (33). However, it is not known whether this is a direct effect of insulin or the result of
248
insulin-mediated normalization of glycemia, leading to reduced glycosuria. Worsening glycemic
249
control resulting in hyperglycemia and glycosuria has been correlated with a decrease in PUA
250
(20, 22). Thus, it is perhaps not surprising that epidemiological studies have shown a decreased
251
risk of UA-related conditions, such as gout, in diabetic compared to non-diabetic individuals –
252
especially in the context of T1D (34). The mechanisms behind the glucose-mediated PUA
253
lowering effects have been explained by osmotic diuresis caused by increased plasma glucose
254
levels (36), proximal tubule alterations (18) or the effect of glucose on renal UA handling (20,
255
36).
256
Our second aim was to determine whether PUA lowering with hyperglycemia in T1D
257
was due to systemic hyperglycemia causing decreased UA production or renal glycosuria
258
causing increased UA excretion. SGLT2 inhibition with empagliflozin under clamped
259
euglycemic conditions was used to increase urinary glucose excretion to determine if glycosuria
260
during euglycemia results in a persistent decrease in PUA through increased renal UA excretion.
11
261
Previous trials with SGLT2 inhibitors in patients with T2D have reported consistent and
262
clinically relevant decreases in PUA levels, however urine UA excretion was not measured and
263
the mechanisms responsible have not, to our knowledge, been clearly elucidated (14, 40). Our
264
post-hoc analysis demonstrated a decline in PUA during euglycemia with glycosuria induced
265
with SGLT2 inhibition, an effect that was accompanied by an increase in UA excretion.
266
Consistent with our observations, in a recent study using healthy controls, SGLT2 inhibition with
267
luseogliflozin resulted in a positive correlation between urine UA and urine glucose excretion
268
(10). The results of the present study provide the first evidence in the T1D population suggesting
269
that hyperglycemia-mediated uricosuria is likely due to renal glycosuria rather than a direct
270
effect of systemic hyperglycemia. PUA lowering effects reported with SGLT2 inhibition may be
271
of clinical relevance, since this may in part explain the potential protective renal and
272
cardiovascular physiological profile that has been linked with this emerging drug class (6).
273
The molecular mechanisms responsible for the uricosuric effect of glucose are not clear.
274
PUA levels depend on the exogenous pool which varies with dietary intake, while the
275
endogenous pool is mainly regulated by hepatic production, intestinal secretion and renal
276
excretion (29). Approximately 70% of UA is excreted into urine, but is easily filtered into the
277
renal tubule and about 90% of filtered UA is reabsorbed by the S1 segment of the proximal
278
convoluted tubule (29). Approximately 10% of filtered UA is excreted (29). Accordingly, our
279
HC showed a FEUA of 6.1±4.1% and T1D during euglycemia 7.3±3.8%. UA reabsorption occurs
280
by intracellular anion transporters on the basolateral membrane – mainly by URAT1 and a more
281
recently discovered GLUT9 isoform 2 (1, 2), and on the apical membrane OAT4 and OAT10 (3,
282
21). Recently, transport experiments in Xenopus oocytes showed that none of the transporters
283
involved in UA reabsorption were influenced by luseogliflozin (10). GLUT9 isoform 2 is a
12
284
facilitative glucose transporter mostly expressed in the kidney and the liver, located on the apical
285
membrane (2). GLUT9 isoform 2 secretes UA in exchange for glucose at 10mM (1).
286
Additionally, GLUT9 isoform 2 is expressed in the collecting ducts where it plays a role in the
287
reabsorption of UA (27). Plasma glucose is mostly filtered in the glomerulus and is concentrated
288
in the proximal tubule. It is possible that during euglycemia the concentration needed for GLUT9
289
stimulation is not reached in the proximal tubule and the lower PUA in T1D during euglycemia
290
vs. HC could occur by mechanisms other than glycosuria- mediated uricosuria. Based on these
291
findings, the results of our study could be explained as follows: glycosuria during SGLT2
292
inhibition stimulates excretion of UA by GLUT9 isoform 2 on the apical membrane of the
293
proximal tubule and possibly inhibits reabsorption of UA in the collecting ducts (Figure 3). Our
294
conclusion may reflect recent in vitro data showing that stimulation of Xenopus oocytes
295
expressing GLUT9 isoform 2 with 10mM D-glucose resulted in UA efflux and stimulation of the
296
oocytes with 100mM D-glucose - thought to be the concentration in the collecting ducts -
297
inhibited the uptake of UA (10). Finally, increased glycosuria and uricosuria could, in the
298
appropriate context, suggest the presence of more generalized, “Fanconi-like” proximal tubular
299
dysfunction. Since SGLT2 inhibition causes minor but statistically significant increases, rather
300
than deceases, in serum potassium, phosphate and bicarbonate, a proximal tubulopathy with this
301
class of agents is very unlikely and has not been reported (38).
302
To examine the functional role of PUA in this otherwise healthy cohort of T1D patients,
303
we correlated PUA with blood pressure and renal hemodynamic function. We found a significant
304
positive correlation between PUA and SBP, negative correlations between PUA and ERPF and
305
PUA and GFR in T1D. In contrast, PUA did not correlate with any of these measures in HC.
306
These observations in T1D patients are consistent with the vasoconstrictive phenotype, as
13
307
suggested by observational studies. For example, the independent association between PUA and
308
blood pressure has been reported in various cohorts, including a subset of the Framingham Heart
309
Study (12, 13). The deleterious effect of PUA on cardiovascular function may be worsened by
310
the hypertensive effect of hyperglycemia in T1D patients (7, 19, 31). Hyperglycemia induces
311
systemic vascular abnormalities such as endothelial dysfunction in humans (19, 31). As a result
312
of the effects of hyperglycemia and neurohormonal activation of the renin angiotensin
313
aldosterone system, UA levels are independently associated with endothelial dysfunction,
314
thereby promoting hypertension, even when UA levels are within the normal range (12, 26).
315
Therefore, lower PUA levels in T1D patients, especially under hyperglycemic conditions, do not
316
necessarily indicate that T1D patients are protected from the deleterious effects of UA. The
317
effects of PUA may be exacerbated by hyperglycemia in T1D patients, leading to exaggerated
318
deleterious hemodynamic consequences despite lower absolute PUA levels. From a clinical
319
perspective, small trials have already started to show that lowering UA exerts anti-proteinuric
320
and antihypertensive effects and could prevent renal functional loss and vascular injury (13, 17,
321
25, 29, 32, 39). Thus, despite the lower UA levels in T1D versus HC, which are further lowered
322
during hyperglycemia, studying UA lowering agents in T1D patients could be a critical step
323
towards preventing progression of diabetes-related complications.
324
Our study has limitations. First, the study cohort consisted of a carefully selected group
325
of patients with uncomplicated disease, limiting the generalizability of the data to populations
326
outside of T1D, or to patients with existing complications. Additionally, although the similar
327
urine urea excretion and thus protein intake suggest that differences in dietary intake of high UA-
328
containing foods were unlikely, consumption of UA was not recorded. Fructose is another
329
exogenous source of UA, which was not recorded in this study, and should be considered in
14
330
future analyses. Finally, while we propose a possible explanation for glycosuria-mediated
331
uricosuria, we could not determine the mechanistic basis at the molecular level. Future studies
332
are needed in order to confirm our hypothesis.
333
In conclusion, glycosuria, rather than the direct effect of hyperglycemia, is responsible
334
for increased uricosuria in T1D patients, and may be mediated by glucose-mediated activation of
335
GLUT9 isoform 2 on the apical membrane of the proximal tubule. Since PUA lowering may lead
336
to renal and vascular protective effects, our data suggest that PUA lowering by SGLT2 inhibition
337
via increased uricosuria may be clinically important. Finally, future studies examining the effect
338
of UA lowering therapies should account for the impact of ambient glucose levels, which cause a
339
clinically relevant uricosuric and consequent PUA lowering effect.
340
15
341
CONTRIBUTIONS:
342
D.Z.I.C, B.A.P., Y.L., P.Y., researched data, wrote the manuscript, contributed to
343
discussion, reviewed/edited manuscript. M.Š., G.K.Y., made contributions to data analysis and
344
interpretation, drafting and editing the manuscript. All authors have approved the final version of
345
this manuscript.
346
347
ACKNOWLEDGEMENTS:
348
This work was supported by Boehringer Ingelheim and Eli Lilly (to D.Z.I.C. and B.A.P.)
349
and operating support from the Canadian Institutes of Health Research and the Heart and Stroke
350
Foundation of Canada. D.Z.I.C. was also supported by a Kidney Foundation of Canada
351
Scholarship and a Canadian Diabetes Association-KRESCENT Program Joint New Investigator
352
Award. The authors would also like to thank Jenny Cheung-Hum for their invaluable assistance
353
with biochemical assays included in this work. Finally, the authors are grateful to the study
354
participants whose time and effort are critical to the success of our research program.
355
356
CONFLICT OF INTEREST:
357
The results presented in this paper have not been published previously in whole or in part.
358
D.Z.I.C. has received speaker honoraria from Boehringer Ingelheim and B.A.P received
359
operational funding with D.Z.I.C. for this work.
360
361
16
362
REFERENCES:
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
1.
Anzai N, Ichida K, Jutabha P, Kimura T, Babu E, Jin CJ, Srivastava S, Kitamura
K, Hisatome I, Endou H, and Sakurai H. Plasma urate level is directly regulated by a voltagedriven urate efflux transporter URATv1 (SLC2A9) in humans. The Journal of biological
chemistry 283: 26834-26838, 2008.
2.
Augustin R, Carayannopoulos MO, Dowd LO, Phay JE, Moley JF, and Moley KH.
Identification and characterization of human glucose transporter-like protein-9 (GLUT9):
alternative splicing alters trafficking. The Journal of biological chemistry 279: 16229-16236,
2004.
3.
Bahn A, Hagos Y, Reuter S, Balen D, Brzica H, Krick W, Burckhardt BC, Sabolic I,
and Burckhardt G. Identification of a new urate and high affinity nicotinate transporter,
hOAT10 (SLC22A13). The Journal of biological chemistry 283: 16332-16341, 2008.
4.
Bonsnes RW, and Dana ES. On the increased uric acid clearance following the
intravenous infusion of hypertonic glucose solutions. The Journal of clinical investigation 25:
386-388, 1946.
5.
Cherney DZ, Miller JA, Scholey JW, Bradley TJ, Slorach C, Curtis JR, Dekker
MG, Nasrallah R, Hebert RL, and Sochett EB. The effect of cyclooxygenase-2 inhibition on
renal hemodynamic function in humans with type 1 diabetes. Diabetes 57: 688-695, 2008.
6.
Cherney DZ, and Perkins BA. Sodium-Glucose Cotransporter 2 Inhibition in Type 1
Diabetes: Simultaneous Glucose Lowering and Renal Protection? Can J Diabetes 2014.
7.
Cherney DZ, Reich HN, Miller JA, Lai V, Zinman B, Dekker MG, Bradley TJ,
Scholey JW, and Sochett EB. Age is a determinant of acute hemodynamic responses to
hyperglycemia and angiotensin II in humans with uncomplicated type 1 diabetes mellitus. Am J
Physiol Regul Integr Comp Physiol 299: R206-214, 2010.
Cherney DZI, Miller JA, Scholey JW, Bradley TJ, Slorach C, Curtis JR, Dekker
8.
MG, Nasrallah R, Hébert RL, and Sochett EB. The effect of cyclooxygenase-2 inhibition on
renal hemodynamic function in humans with type 1 diabetes. Diabetes 57: 688-695, 2008.
9.
Cherney DZI, Perkins BA, Soleymanlou N, Maione M, Lai V, Lee A, Fagan NM,
Woerle HJ, Johansen OE, Broedl UC, and von Eynatten M. Renal hemodynamic effect of
sodium-glucose cotransporter 2 inhibition in patients with type 1 diabetes mellitus. Circulation
129: 587-597, 2014.
10.
Chino Y, Samukawa Y, Sakai S, Nakai Y, Yamaguchi JI, Nakanishi T, and Tamai I.
SGLT2 inhibitor lowers serum uric acid through alteration of uric acid transport activity in renal
tubule by increased glycosuria. Biopharmaceutics & drug disposition 2014.
11.
Choi HK, and Ford ES. Haemoglobin A1c, fasting glucose, serum C-peptide and insulin
resistance in relation to serum uric acid levels--the Third National Health and Nutrition
Examination Survey. Rheumatology (Oxford, England) 47: 713-717, 2008.
12.
Erdogan D, Gullu H, Caliskan M, Yildirim E, Bilgi M, Ulus T, Sezgin N, and
Muderrisoglu H. Relationship of serum uric acid to measures of endothelial function and
atherosclerosis in healthy adults. Int J Clin Pract 59: 1276-1282, 2005.
13.
Feig DI, Rodriguez-Iturbe B, Nakagawa T, and Johnson RJ. Nephron number, uric
acid, and renal microvascular disease in the pathogenesis of essential hypertension. Hypertension
48: 25-26, 2006.
14.
Ferrannini E, Seman L, Seewaldt-Becker E, Hantel S, Pinnetti S, and Woerle HJ. A
Phase IIb, randomized, placebo-controlled study of the SGLT2 inhibitor empagliflozin in
patients with type 2 diabetes. Diabetes, obesity & metabolism 15: 721-728, 2013.
17
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
15.
Ficociello LH, Rosolowsky ET, Niewczas MA, Maselli NJ, Weinberg JM,
Aschengrau A, Eckfeldt JH, Stanton RC, Galecki AT, Doria A, Warram JH, and
Krolewski AS. High-normal serum uric acid increases risk of early progressive renal function
loss in type 1 diabetes: results of a 6-year follow-up. Diabetes care 33: 1337-1343, 2010.
16.
Florijn KW, Barendregt JN, Lentjes EG, van Dam W, Prodjosudjadi W, van Saase
JL, van Es LA, and Chang PC. Glomerular filtration rate measurement by "single-shot"
injection of inulin. Kidney international 46: 252-259, 1994.
17.
Goicoechea M, de Vinuesa SG, Verdalles U, Ruiz-Caro C, Ampuero J, Rincon A,
Arroyo D, and Luno J. Effect of allopurinol in chronic kidney disease progression and
cardiovascular risk. Clin J Am Soc Nephrol 5: 1388-1393, 2010.
18.
Gonzalez-Sicilia L, Garcia-Estan J, Martinez-Blazquez A, Fernandez-Pardo J,
Quiles JL, and Hernandez J. Renal metabolism of uric acid in type I insulin-dependent diabetic
patients: relation to metabolic compensation. Hormone and metabolic research = Hormon- und
Stoffwechselforschung = Hormones et metabolisme 29: 520-523, 1997.
19.
Gordin D, Ronnback M, Forsblom C, Makinen V, Saraheimo M, and Groop PH.
Glucose variability, blood pressure and arterial stiffness in type 1 diabetes. Diabetes Res Clin
Pract 80: e4-7, 2008.
20.
Gotfredsen A, McNair P, Christiansen C, and Transbol I. Renal hypouricaemia in
insulin treated diabetes mellitus. Clinica chimica acta; international journal of clinical chemistry
120: 355-361, 1982.
21.
Hagos Y, Stein D, Ugele B, Burckhardt G, and Bahn A. Human renal organic anion
transporter 4 operates as an asymmetric urate transporter. Journal of the American Society of
Nephrology : JASN 18: 430-439, 2007.
22.
Herman JB, and Goldbourt U. Uric acid and diabetes: observations in a population
study. Lancet 2: 240-243, 1982.
23.
Jalal DI, Rivard CJ, Johnson RJ, Maahs DM, McFann K, Rewers M, and SnellBergeon JK. Serum uric acid levels predict the development of albuminuria over 6 years in
patients with type 1 diabetes: findings from the Coronary Artery Calcification in Type 1 Diabetes
study. Nephrology, dialysis, transplantation : official publication of the European Dialysis and
Transplant Association - European Renal Association 25: 1865-1869, 2010.
24.
Jones SL, Kontessis P, Wiseman M, Dodds R, Bognetti E, Pinto J, and Viberti G.
Protein intake and blood glucose as modulators of GFR in hyperfiltering diabetic patients.
Kidney international 41: 1620-1628, 1992.
Kanbay M, Huddam B, Azak A, Solak Y, Kadioglu GK, Kirbas I, Duranay M,
25.
Covic A, and Johnson RJ. A randomized study of allopurinol on endothelial function and
estimated glomular filtration rate in asymptomatic hyperuricemic subjects with normal renal
function. Clin J Am Soc Nephrol 6: 1887-1894, 2011.
26.
Kanbay M, Yilmaz MI, Sonmez A, Turgut F, Saglam M, Cakir E, Yenicesu M,
Covic A, Jalal D, and Johnson RJ. Serum uric acid level and endothelial dysfunction in
patients with nondiabetic chronic kidney disease. Am J Nephrol 33: 298-304, 2011.
27.
Kimura T, Takahashi M, Yan K, and Sakurai H. Expression of SLC2A9 isoforms in
the kidney and their localization in polarized epithelial cells. PloS one 9: e84996, 2014.
28.
Lewis G, and Maxwell AP. Risk factor control is key in diabetic nephropathy. The
Practitioner 258: 13-17, 12, 2014.
29.
Lytvyn Y, Perkins BA, and Cherney DZI. Uric Acid as a Biomarker and a Therapeutic
Target in Diabetes. Canadian Journal of Diabetes In Press: 2014.
18
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
30.
Maahs DM, Caramori L, Cherney DZ, Galecki AT, Gao C, Jalal D, Perkins BA,
Pop-Busui R, Rossing P, Mauer M, and Doria A. Uric Acid Lowering to Prevent Kidney
Function Loss in Diabetes: The Preventing Early Renal Function Loss (PERL) Allopurinol
Study. Current diabetes reports 13: 550-559, 2013.
31.
Miller JA. Impact of hyperglycemia on the renin angiotensin system in early human type
1 diabetes mellitus. Journal of the American Society of Nephrology : JASN 10: 1778-1785, 1999.
32.
Momeni A, Shahidi S, Seirafian S, Taheri S, and Kheiri S. Effect of allopurinol in
decreasing proteinuria in type 2 diabetic patients. Iran J Kidney Dis 4: 128-132, 2010.
33.
Quinones Galvan A, Natali A, Baldi S, Frascerra S, Sanna G, Ciociaro D, and
Ferrannini E. Effect of insulin on uric acid excretion in humans. The American journal of
physiology 268: E1-5, 1995.
34.
Rodriguez G, Soriano LC, and Choi HK. Impact of diabetes against the future risk of
developing gout. Annals of the rheumatic diseases 69: 2090-2094, 2010.
35.
Rosolowsky ET, Ficociello LH, Maselli NJ, Niewczas MA, Binns AL, Roshan B,
Warram JH, and Krolewski AS. High-normal serum uric acid is associated with impaired
glomerular filtration rate in nonproteinuric patients with type 1 diabetes. Clin J Am Soc Nephrol
3: 706-713, 2008.
36.
Skeith MD, Healey LA, and Cutler RE. Effect of phloridzin on uric acid excretion in
man. The American journal of physiology 219: 1080-1082, 1970.
37.
Wang C, Pan Y, Zhang QY, Wang FM, and Kong LD. Quercetin and allopurinol
ameliorate kidney injury in STZ-treated rats with regulation of renal NLRP3 inflammasome
activation and lipid accumulation. PloS one 7: e38285, 2012.
38.
Weir MR, Kline I, Xie J, Edwards R, and Usiskin K. Effect of canagliflozin on serum
electrolytes in patients with type 2 diabetes in relation to estimated glomerular filtration rate
(eGFR). Current medical research and opinion 30: 1759-1768, 2014.
39.
Whelton A, Macdonald PA, Zhao L, Hunt B, and Gunawardhana L. Renal function
in gout: long-term treatment effects of febuxostat. Journal of clinical rheumatology : practical
reports on rheumatic & musculoskeletal diseases 17: 7-13, 2011.
40.
Wilding JP, Ferrannini E, Fonseca VA, Wilpshaar W, Dhanjal P, and Houzer A.
Efficacy and safety of ipragliflozin in patients with type 2 diabetes inadequately controlled on
metformin: a dose-finding study. Diabetes, obesity & metabolism 15: 403-409, 2013.
Yang GK, Maahs DM, Perkins B, and Cherney DZI. Renal Hyperfiltration and
41.
Systemic Blood Pressure in Patients with Uncomplicated Type 1 Diabetes Mellitus. PloS one 8:
e68908, 2013.
488
489
490
491
492
493
494
495
496
497
498
499
19
500
501
502
503
504
505
506
507
508
509
510
Table 1. Baseline subject characteristics and UA, sodium, glucose handling in HC and
patients with T1D during euglycemia and hyperglycemia (mean ± standard deviation).
24 hour protein intake: estimated by the formula ([urine urea X 0.18] + 14) / weight in kg. Values are
mean ± standard deviation. *p<0.05 for HC vs. T1D. †p<0.05 when comparing parameters of T1D
between hyperglycemia and euglycemia states. FEUA: fractional excretion of UA; FENa: fractional
excretion of sodium; HR: heart rate; SBP: systolic blood pressure; DBP: diastolic blood pressure;
ERPF: effective renal plasma flow; GFR: glomerular filtration rate; RBF: renal blood flow; RVR: renal
vascular resistance; PRA: plasma renin activity; HC: healthy controls; T1D: type 1 diabetic patients.
511
512
513
Figure 1. PUA (A), FEUA (B) and urine glucose/creatinine (C) levels in HC (n=41), and T1D
(n=66) during euglycemic (EU) and hyperglycemic (HYP) conditions. The bars above
cohorts in each figure represent significance levels of p<0.05
514
515
516
517
Figure 2. PUA (A), FEUA (B) and urine glucose/creatinine (C) levels in T1D (n=40) during
euglycemic conditions at baseline and after treatment with the SGLT2 inhibitor
empagliflozin (EMPA). The bars above cohorts in each figure represent significance levels of
p<0.05
518
519
520
521
Figure 3. Proposed hypothesis for glycosuria mediated hyperuricemia in T1D patients,
supported by findings in this study and by Chino et al.(10). SGLT2i: Sodium glucose
transporter 2 inhibitor; GLUT9: Glucose transporter 9; URAT1: Urate Transporter 1; UA: Uric
Acid.
522
523
524
525
Figure 4. Linear regression analysis of PUA with SBP in T1D during euglycemia (A), with
ERPF in T1D during euglycemia (B), with GFR in T1D during hyperglycemia (C). T1D
n=40; SBP: systolic blood pressure; ERPF: effective renal plasma flow; GFR: glomerular
filtration rate.
526
527
528
529
530
531
532
533
534
535
536
20
Table 1. Baseline subject characteristics and UA, sodium, glucose handling in HC and
patients with T1D during euglycemia and hyperglycemia (mean ± standard deviation).
Parameter
HC (n=41)
T1D (n=66)
Euglycemia
Hyperglycemia
Baseline parameters
Males
Age (years)
Diabetes duration (years)
Weight (kg)
Height (m)
Body mass index (kg/m2)
Hemoglobin A1c - mmol/mol (%)
19 (43%)
28.4±7.1
70.4±11.8
1.74±0.09
23.3±3.0
34.8±3.6 (5.3±0.3)
35 (53%)
25.0±6.0
17.0±6.6
73.9±13.7
1.73±0.09
24.8±3.9
66.7±16.1 (8.2±1.5) *
-
24 hour urine sodium (mmol/day)
24 hour protein intake (g/kg/day)
177±61
1.1±0.3
169±85
1.0±0.3
-
305±75
248±170
6.1±4.1
0.84±0.60
0.02±0.03
228±62*
257±121
7.3±3.8
0.87±0.56
1.4±3.2*
199±65†
339±161 †
11.6±6.7 †
1.63±0.89†
9.8±10.4 †
60±9
112±12
67±8
74±13*
115±10
66±6
72±11†
117±11†
66±8
Renal hemodynamic function
ERPF (mL/min/1.73m2)
GFR (mL/min/1.73m2)
Filtration fraction
RBF (mL/min/1.73m2)
RVR (mmHg/L/min)
653±157
116±12
0.19±0.04
1063±259
0.081±0.020
824±276*
147±40*
0.19±0.06
1310±434*
0.069±0.021*
853±253
159±39†
0.19±0.05
1305±419
0.055±0.016†
Circulating neurohormones
Aldosterone (ng/dL)
PRA (ng/mL/h)
Angiotensinogen (ng/mL)
Angiotensin II
245±254
1.34±1.14
1264±1000
11.6±8.4
45±31*
0.53±0.43*
1092±722
3.1±3.5*
31±10 †
0.35±0.27†
1076±742
2.2±2.2
Sodium, glucose, uric acid handling
Plasma uric acid (µmol/L)
Urine uric acid/creatinine ratio
FEUA (%)
FENa (%)
Urine glucose/creatinine ratio
Systemic hemodynamic function
HR (beats per minute)
SBP (mmHg)
DBP (mmHg)
24 hour protein intake: estimated by the formula ([urine urea X 0.18] + 14) / weight in kg. Values are
mean ± standard deviation. *p<0.05 for HC vs. T1D. †p<0.05 when comparing parameters of T1D
between hyperglycemia and euglycemia states. FEUA: fractional excretion of UA; FENa: fractional
excretion of sodium; HR: heart rate; SBP: systolic blood pressure; DBP: diastolic blood pressure;
ERPF: effective renal plasma flow; GFR: glomerular filtration rate; RBF: renal blood flow; RVR: renal
vascular resistance; PRA: plasma renin activity; HC: healthy controls; T1D: type 1 diabetic patients.
Figure 1. PUA (A), FEUA (B) and urine glucose/creatinine (C) levels in HC (n=41), and T1D
(n=66) during euglycemic (EU) and hyperglycemic (HYP) conditions.
B
Y
D
D
T
T
1
1
T
D
1
T
-
T
1
1
H
D
Y
-
P
E
E
U
C
H
D
0
R
R
P
-
T
1
H
D
Y
-
H
E
E
U
C
0
20
R
0
E
5
40
H
100
P
10
60
-
U A
200
U
(% )
15
80
C
300
E
20
F E
P U A ( P m o l/L )
U r in e G lu c o s e /C r e a tin in e
C
400
H
A
The bars above cohorts in each figure represent significance levels of p<0.05
Figure 2. PUA (A), FEUA (B) and urine glucose/creatinine (C) levels in T1D (n=40) during
euglycemic conditions at baseline and after treatment with the SGLT2 inhibitor
empagliflozin (EMPA).
300
15
U A
200
U r in e G lu c o s e /C r e a tin in e
20
(% )
400
10
F E
100
5
0
0
80
60
40
20
0
A
B
A
B
The bars above cohorts in each figure represent significance levels of p<0.05
1
A
P
E
M
IN
S
E
M
E
L
E
S
E
L
P
A
E
P
M
IN
L
B
A
S
E
IN
A
E
E
P U A ( P m o l/L )
C
B
A
Figure 3. Proposed hypothesis for glycosuria mediated hyperuricemia in T1D patients,
supported by findings in this study and by Chino et al.(10).
SGLT2i: Sodium glucose transporter 2 inhibitor; GLUT9: Glucose transporter 9; URAT1: Urate
Transporter 1; UA: Uric Acid.
Figure 4. Linear regression analysis of PUA with SBP in T1D during euglycemia (A), with
ERPF in T1D during euglycemia (B), with GFR in T1D during hyperglycemia (C).
B
C
400
r= 0 .2 7
300
200
100
0
80
100
120
140
S B P (m m H g )
160
500
r= -0 .2 7
p = 0 .0 2 9
P U A ( µ m o l/L )
P U A ( µ m o l/L )
400
p = 0 .0 3 1
300
200
100
0
P U A ( µ m o l/L )
A
r= -0 .2 7
p = 0 .0 2 8
400
300
200
100
0
0
500
1000
1500
E R P F ( m L /m in /1 .7 3 m 2 )
2000
0
100
200
300
G F R ( m L /m in /1 .7 3 m 2 )
T1D n=40; SBP: systolic blood pressure; ERPF: effective renal plasma flow; GFR: glomerular
filtration rate.
2