Standardized, Systemic Phenotypic Analysis of Umod C93F and Umod A227T Mutant Mice

Uromodulin-associated kidney disease (UAKD) summarizes different clinical features of an autosomal dominant heritable disease syndrome in humans with a proven uromodulin (UMOD) mutation involved. It is often characterized by hyperuricemia, gout, alteration of urine concentrating ability, as well as a variable rate of disease progression inconstantly leading to renal failure and histological alterations of the kidneys. We recently established the two Umod mutant mouse lines Umod C93F and Umod A227T on the C3H inbred genetic background both showing kidney defects analogous to those found in human UAKD patients. In addition, disease symptoms were revealed that were not yet described in other published mouse models of UAKD. To examine if further organ systems and/or metabolic pathways are affected by Umod mutations as primary or secondary effects, we describe a standardized, systemic phenotypic analysis of the two mutant mouse lines Umod A227T and Umod C93F in the German Mouse Clinic. Different genotypes as well as different ages were tested. Beside the already published changes in body weight, body composition and bone metabolism, the influence of the Umod mutation on energy metabolism was confirmed. Hematological analysis revealed a moderate microcytic and erythropenic anemia in older Umod mutant mice. Data of the other analyses in 7-10 month-old mutant mice showed single small additional effects.


Introduction
The uromodulin gene (UMOD) encodes the uromodulin glycoprotein which is the most abundant urinary protein in mammals. It is synthesized exclusively and abundantly in the cells of the thick ascending limb of Henle's loop (TALH) without macula densa cells. Once synthesized, the uromodulin precursor is processed within the endoplasmic reticulum (ER) and Golgi complex into mature glycoprotein, transported to the luminal membrane and excreted with urine.
In humans, usually heterozygous UMOD mutations result in a dominant heritable disease syndrome which is summarized as uromodulin-associated kidney disease (UAKD) or uromodulin storage disease. The disease is characterized by TALH dysfunction due to disturbed mutant uromodulin trafficking that in consequence leads to reduced fractional excretion of uric acid, leading in most cases to gout, mild defects of urine concentrating ability, as well as a variable rate of disease progression inconstantly leading to renal failure. Histological alterations of the kidneys include tubulointerstitial fibrosis with moderate inflammatory cell infiltrates and cysts [1,2].
To date, various Umod mutant mouse models are published. Two different Umod functional knockout models were established on the genetic background of 129/Sv and/or C57BL/6 inbred mice [3,4]. Compared to wild-type mice, homozygous knockout mice of both lines showed normal development and growth as well as behavior and fertility. The histological kidney structures also were not different, and they exhibited only minor changes in steady-state renal function. This indicates that loss of uromodulin expression is not sufficient to cause UAKD [5][6][7].
In addition, two transgenic mouse lines were published. The first transgenic line expressed C148W human mutant uromodulin -which is an UMOD mutation known to cause UAKD in humans -under the control of the mouse Umod promoter in C57BL/6 inbred mice. Uromodulin accumulation was observed in TALH cells of transgenic mice, but mouse UMOD urinary excretion did not decrease and no clinical renal phenotype was observed also in aged mice [8]. The second transgenic mice expressing the corresponding C147W murine mutant uromodulin in the genetic background of FVB/N inbred mice showed the same body weight as controls. They exhibited a clinical phenotype with features of strong UAKD and renal failure with tubular necrosis at an age of 6 months [9].
In the phenotype-based Munich ENU mouse mutagenesis project using C3HeB/FeJ (C3H) inbred mice as genetic background, several mutant lines were established showing increased plasma urea levels as a parameter indicative of kidney disease [10]. Two of the lines showed dominant Umod mutations, i.e. lines Umod C93F and Umod A227T . Similar to UAKD in humans, they exhibit a phenotype of impaired kidney function indicated by increased plasma urea and creatinine values. The functional alterations of both mutants indicated a gain-of-toxic function of mutant uromodulin, leading to TALH dysfunction due to disturbed uromodulin trafficking. Onset, severity and progression of UAKD symptoms in these mice were influenced by the type of Umod mutation and the allelic status. Thus, the severity of the kidney alterations was similar in Umod C93F heterozygous mutants and Umod A227T homozygous mutants. In addition, we found phenotypic alterations previously not described in mutants for this gene. Changes in body weight, body composition and bone metabolism of lines Umod C93F and Umod A227T [11,12] as well as an influence on energy metabolism in line Umod A227T [12] were revealed.
Adult homozygous mutants of both genders of line Umod A227T had a significantly lower body weight than wild-type mice. Heterozygous mutants showed an intermediate state. Analysis of body composition demonstrated that fat mass and fat content were significantly decreased in mutants whereas lean content was increased. Increased plasma calcium and alkaline phosphatase (ALP) activity were found in the mutants. To elucidate the long-term effect of hypercalciuria, analysis of the skeleton showed that bone mineral density and bone mineral content were significantly decreased in 4-month-old Umod A227T mutant animals, indicating osteopenia [12]. Analogous results with even more pronounced alterations were revealed in 9month-old heterozygous mutants of line Umod C93F [11].
Analysis of energy metabolism in line Umod A227T at the age of 3 months revealed that body mass and body temperature were reduced in mutant mice. Metabolic rate was decreased in mutants as expected for lower body mass but food intake was significantly increased [12].
In total, we already revealed pathological effects in other organ systems than the kidneys, i.e. skeletal and metabolic alterations in Umod mutant mouse lines. To examine if further organ systems and/or metabolic pathways are affected by Umod mutations as primary or secondary effects, we describe a standardized, systemic phenotypic analysis of the two mutant mouse lines Umod A227T and Umod C93F in the German Mouse Clinic (http://www.mouseclinic.de).

Materials and Methods
Both dominant mutant lines Umod C93F and Umod A227T were established in the Munich ENU mouse mutagenesis project using C3HeB/FeJ (C3H) inbred mice as genetic background [13]. Maintenance of the lines comprised the repeated backcross to C3H wild-type mice leading to the subsequent loss of essentially all non-causative ENU mutations that were not linked to the Umod mutation. The systemic, comprehensive phenotypic analysis was carried out in the German Mouse Clinic at the Helmholtz Zentrum München by using standardized examination protocols (http:// www.mouseclinic.de). The analysis covers over 300 parameters in the areas of allergy, behavior, bone and cartilage, cardiovascular analysis, clinical chemistry, energy metabolism, eye analysis and vision, immunology, lung function, molecular phenotyping, neurology, nociception, pathology, and steroid metabolism. The complete protocols of the examinations are described under http:// www.mouseclinic.de [14][15][16]. The standard workflow of the primary phenotypic analysis in the German Mouse Clinic was carried out.
First, the primary standard analysis of Umod A227T heterozygous mutant and homozygous mutant mice was carried out at an age of 2-4 months. After that, the primary standard analysis of Umod C93F heterozygous mutants was carried out at an age of 7-10 months also to especially reveal long-term secondary effects of Umod mutations (Table S1). For line Umod A227T , 10 homozygous mutant, 10 heterozygous mutant, and 10 wild-type control littermates of each sex were used. The mice analyzed were a fifth generation backcross of the original ENU-mutated founder mouse to C3H wild-type mice. For line Umod C93F , 41 heterozygous mutants and 39 wildtype control littermates were used by dividing them in two groups for two parallel phenotype analyses. The line was analyzed after backcrossing the original ENU-mutated founder mouse for more than 10 generations to C3H wild-type mice. The analysis of both mutations took place in the German Mouse Clinic at different time periods. The number of animals analyzed for both mutations was 8-11 animals per sex and genotype (except of otherwise stated in the text of the respective Results section).
If not otherwise stated, statistical analysis of data was carried out by Student's t-test. Data are shown as mean ± standard error of the mean. Significant differences are indicated for P < 0.05, 0.01, and 0.001. In addition, the data of the feeding efficiency analysis were examined by using two-way repeated measures ANOVA.

Ethics statement
Mouse husbandry and all tests were carried out under the approval of the responsible animal welfare authority (Regierung von Oberbayern, Germany).

Results
Phenotypic analysis of mutant mice in the German Mouse Clinic aims at collecting and delivering with free access comprehensive phenome data of a high number of mutant mouse lines in a standardized manner. The phenotype reports of both mutant lines Umod A227T (see line "HST012") and Umod C93F (see line "HST001") are deposited online (http:// 146.107.35.38/phenomap/jsp/annotation/public/phenomap.jsf). First, the primary standard analysis of Umod A227T heterozygous mutant and homozygous mutant mice was carried out at an age of 2-4 months. As Umod A227T homozygous mutants and Umod C93F heterozygous mutants at the same age have a similar severity of the kidney alterations [11], the primary standard analysis of Umod C93F heterozygous mutants was carried out at an age of 7-10 months (Table S1).

Additional clinical chemistry analysis
In addition to the previously published results of the clinical chemical parameters of blood plasma and urine of both Umod mutant lines, hematological parameters of line Umod A227T were repeatedly tested at the age of 16 and 19 weeks. Homozygous mutants and heterozygous mutants of both sexes showed small differences in hematology parameters, i.e. decreased hemoglobin, hematocrit and mean corpuscular volume at both time points versus wild-type controls ( Table 1, and data not shown for the second measurement). Blood gas analysis tested at the age of 9 months indicated the absence of obvious alterations ( Table 2).
In line Umod C93F , 35-and 43-week-old mice were used. The hematological analyses reproducibly revealed a moderate microcytic and erythropenic anemia in heterozygous mutants indicated by decreased values for red blood cell count, hemoglobin, hematocrit, mean corpuscular volume and mean corpuscular hemoglobin content (Table 1, and data not shown for the second measurement). Additional hematological analyses in Umod C93F homozygous mutants, heterozygous mutants and wild-type controls at the earlier age of 12 weeks revealed the onset of the first alterations in the hematological parameters (not shown).
In addition, the intraperitoneal glucose tolerance test (IpGTT) was carried out in line Umod C93F at the age of 8 months (n=10 per genotype and sex). Both male (P < 0.05) and female (P = 0.08) heterozygous mutants showed lower fasted plasma glucose levels than the wild-type controls (median [25% level,  mutant mice versus wild-type controls with normal body fat content. Urine analysis at the age of 9 months (n = 8-10 per genotype and sex) may indicate an increased urinary glucose excretion per gram body weight in the heterozygous mutants versus controls (excretion/24 h/25 g body weight (mean ± SD): 325 ± 131 µg vs. 227 ± 106 µg in males (P = 0.133), and 369 ± 64 µg vs. 283 ± 61 µg in females (P = 0.009)). No significant difference of urinary glucose excretion relative to the body weight was observed in 13-16 weeks-old Umod A227T homozygous mutants and heterozygous mutants compared to wild-type controls [12].

Additional analysis of dysmorphology
In addition to the previously published differences in body weight and body composition as well as bone metabolism, systemic morphological investigation via visual inspection and X-ray analysis according to standardized protocols were carried out in both lines. In line Umod A227T , visual inspection at the age of 10-12 weeks and X-ray analysis of 4-month-old animals revealed no genotype-specific differences between homozygous mutants, heterozygous mutants and wild-type controls. The same was true for the hearing ability in 2-monthold mice examined by the clickbox test using a sound of 20 kHz. Analogous results were observed in 31-36 week-old mice of line Umod C93F .

Additional analysis of energy metabolism
The alterations of the energy metabolism in 3-month-old mutants of line Umod A227T have been already published [12]. The respective plasma lipid (cholesterol, triglycerides) values were decreased in repeated analyses of homozygous mutant and heterozygous mutant mice of line Umod A227T (at 16 and 19 weeks of age) and of heterozygous mutant mice of line Umod C93F (at 35 and 43 weeks of age) compared to wild-type controls [11,12]. In contrast, in line Umod C93F at 7 weeks of age and 12 weeks of age (except of triglycerides in males), homozygous mutants with already decreased body weight showed higher plasma lipid (cholesterol, triglycerides) values compared to heterozygous mutants and/or wild-type controls. Umod C93F heterozygous mutants and wild-type controls at 84-99 weeks of age showed inconsistent results for both parameters (data not shown).
In line Umod C93F , two independent analyses were carried out with two different groups of mice (n=5-7 per genotype and sex) at different time points. First, 37-week-old mice were examined both during ad libitum feeding and under food restriction conditions with the feeding efficiency protocol that was used earlier in the German Mouse Clinic. Secondly, indirect calorimetry as the common screen of energy metabolism in the German Mouse Clinic was done on 33-week-old mice under ad libitum conditions.
In the indirect calorimetry analysis (Table 3), heterozygous mutants showed reduced body mass as well as a clear reduction in body temperature by 0.9°C in males and by 0.7°C in females (P = 0.55). Food intake (not corrected for spillage) during 21 hours of gas exchange measurement was higher in heterozygous mutants. Both mean and minimum oxygen consumption (VO 2 ) were reduced in heterozygous mutant mice. As VO 2 mainly depends on body surface, size or mass, a linear model was calculated including body mass as independent factor. According to the model, both male and female heterozygous mutants showed a reduction in oxygen consumption by about 11% when adjusted for body mass (not shown). The mean respiratory quotient remained unaltered in heterozygous mutants and wild-type controls.
Having used the feeding efficiency protocol with 5-7 mice per genotype and sex (Table 3), the data were examined by using two-way repeated measures ANOVA to analyze interactions between the two factors of genotype and feeding on the parameters of body weight (P = 0.076 in males, and P < 0.01 in females), rectal body temperature (P = 0.051 in males, and P < 0.01 in females), energy content of feces (P > 0.05 in males and in females), metabolized energy (P > 0.05 in males and in females), and assimilation coefficient (P < 0.01 in males, and P < 0.05 in females).
Heterozygous mutants also showed reduced body mass and body temperature both under ad libitum (for 7 days) and fasting (for 7 days with 60% of ad libitum consumption) conditions. During fasting no obvious difference in body mass loss between heterozygous mutant and control mice was detected compared to the ad libitum period. However, reduction in body temperature was about 0.5°C in wild-type controls compared to the ad libitum period, but much stronger in heterozygous mutants which entered states of hypothermia with a body temperature being 2-3°C below that of the controls. Another significant difference during fasting was revealed regarding energy assimilation efficiency which increased in heterozygous mutants compared to the wild-type controls. During fasting, energy content of feces increased in both genotypes compared to the respective values of the ad libitum period which is usually observed in this test (Table 3).

Cardiovascular analysis
Non-invasive blood pressure analysis in conscious mice was carried out to determine pulse, systolic and diastolic blood pressure as well as mean arterial pressure (MAP) ( Table 4). In addition, plasma concentration of the N-terminal fragment of the pro-atrial natriuretic peptide (Nt-proANP) was analyzed using the ELISA technique (Biomedica Medizinprodukte, Austria). ANP is a cardiac hormone predominantly secreted by atrial myocytes in response to cardiac filling pressures.
In line Umod A227T , homozygous mutant, heterozygous mutant and wild-type control mice at an age of 12-15 weeks showed no genotype-specific differences in the blood pressure. The range of the means was 496-497 beats per minute (bpm) in males and 526-538 bpm in females for the pulse, 104-108 mm Hg in males and 106-113 mm Hg in females for the systolic blood pressure, 94-97 mm Hg in males and 96-105 mm Hg in females for the diastolic blood pressure, and 97-100 mm Hg in males and 99-107 mm Hg in females for the mean arterial pressure (MAP). Plasma Nt-proANP tended to be increased in homozygous mutants vs. wild-type controls (means ± SD: 2.64 ± 0.60 vs. 1.87 ± 1.03 nmol/l in males and 1.97 ± 0.35 vs. 1.54 ± 0.51 nmol/l in females (P < 0.05)). Heterozygous mutants showed intermediate means [12].
In line Umod C93F , 32-38-week-old mice were used. 32-weekold heterozygous mutant and wild-type mice showed no genotype-specific differences in the blood pressure. The range of the means was 558-574 bpm in males and 572-579 bpm in females for the pulse, 99-108 mm Hg in males and 106-107 mm Hg in females for the systolic blood pressure, 88-96 mm Hg in males and 95-96 mm Hg in females for the diastolic blood pressure, and 92-99 mm Hg in males and 99 mm Hg in females for the mean arterial pressure (MAP). Plasma Nt-proANP of 38-week-old mice tended to be increased in heterozygous mutants vs. wild-type controls (1.21 ± 0.10 vs. 0.80 ± 0.07 nmol/l in males and 2.20 ± 0.27 vs. 1.73 ± 0.22 nmol/l in females).
In addition, echocardiography was done in 36-week-old mice of line Umod C93F to analyze the left ventricular function in contraction (left ventricle inner diameter (LVID) examined in systole and diastole) and pump capacity (fractional shortening, ejection fraction and left ventricular volume). Decreased endsystolic and end-diastolic left ventricular diameters were found in heterozygous mutants which is most likely related to the overall lower body weight of the heterozygous mutants. Comparing heterozygous mutants and wild-type controls, parameters representing the contractile function of the left ventricle, i.e. fractional shortening (35.0 ± 3.4% vs. 33.5 ± 2.5% in males, and 38.2 ± 1.5% vs. 37.4 ± 2.2% in females) and ejection fraction (63.9 ± 4.8% vs. 62.0 ± 3.5% in males, and 69.2 ± 1.9% vs. 67.5 ± 3.0% in females) were not different (Table 4). Two independent analyses (indirect calorimetry as standard screen in the German Mouse Clinic at 33 weeks of age; feeding efficiency protocol at 37 weeks of age both under ad libitum (for 7 days) and fasting (for 7 days with 60% of ad libitum consumption) conditions) were carried out with two independent groups of mice (n = 5-7 per genotype and sex). Data are presented as mean ± standard error of mean. Student's t-test vs. wild-type controls: a P < 0.05, b P < 0.01, c P < 0.001.

Immunology
Peripheral blood leukocytes were isolated and CD45 + viable cells were subsequently analyzed for the identification of main lineages (T cells, B cells, granulocytes, NK cells, monocytes) und subpopulations. Plasma antibody levels were determined simultaneously in the same samples with monoclonal antimouse antibodies conjugated to beads of different regions (Biorad, USA) ( Table 5). The presence of rheumatoid factor and anti-DNA antibodies was evaluated by indirect ELISA with rabbit IgG (Sigma-Aldrich, Germany) and calf thymus DNA (Sigma-Aldrich), respectively, as antigens and AP-conjugated goat anti-mouse secondary antibody (Sigma-Aldrich).
In line Umod A227T , 14-week-old mice were tested. No obvious differences in the frequencies of main leukocyte populations between mutant mice compared to wild-type controls appeared. A statistically higher expression of CD62L within the CD8 + cell cluster occurred in female mutants (homozygous mutants vs. heterozygous mutants vs. controls: CD4 + /CD62L + (%): 23.8 ± 2.4 vs. 21.2 ± 2.1 vs. 17.8 ± 2.6 in males (not significant), and 18.6 ± 2.5 vs. 20.0 ± 2.4 (P < 0.05) vs. 12.3 ± 2.0 in females; CD8 + /CD62L + (%): 33.6 ± 1.9 vs. 33.2 ± 3.1 vs. 28.2 ± 3.0 in males (not significant), and 39.8 ± 3.6 (P < 0.05) vs. 39.0 ± 2.8 (P < 0.05) vs. 29.5 ± 3.1 in females). CD62L is expressed on naive T cells, as well as central memory T cells. Loss of CD62L (shedding) can occur during blood sample preparation due to release of NAD and ATP from lysed erythrocytes, and can vary substantially between inbred mouse strains. Analysis of blood plasma revealed very high antibody levels in all male mice, both mutants and controls, but no differences in the levels of antibodies in the mutants versus controls. Also no changes in autoantibodies (anti-DNA antibodies, rheumatoid factor) were observed (not shown).

Lung function
Analysis of lung function was not carried out in line Umod A227T . In line Umod C93F , spontaneous breathing patterns during sleep, rest and activity were analyzed in six mice per genotype and sex at 38 weeks of age by whole body plethysmography ( Table 6). The mean of all breathing frequencies (mean f) measured during the 40-minute examination period was calculated as a general parameter to assess whether the duration of rest and activity was similar in all groups. It did not differ between heterozygous mutants and wild-type controls. Specific tidal volumes and specific minute ventilations (sTV and sMV) during sleep, rest and activity, respectively, were calculated by relating the absolute values to the body weight of the animals. The values were comparable between heterozygous mutants and controls; during activity, female heterozygous mutants showed even higher sTV and sMV. Overall, observed differences were within the physiological ranges and of minor relevance. Therefore it was suggested that the mutation Umod C93F does not affect the respiratory system.

Neurology
Analysis of the basic neurological functions was not carried out in line Umod A227T . In line Umod C93F , 30-week-old mice were tested by modified SHIRPA, grip strength and rotarod analysis. The modified SHIRPA protocol is a semi-quantitative screening method for the overall qualitative analysis of abnormal phenotypes in mice and includes 23 test parameters each contributing to the overall assessment of general health, posture and movement, autonomic reflexes, as well as behavioral aspects. Observation of undisturbed behavior was done in a glass cylinder (11 cm in diameter). The mice were then transferred to an arena consisting of a clear Perspex box (42 × 26 × 18 cm) in which a Perspex sheet on the floor is marked with 15 squares. Locomotor activity was similar in heterozygous mutants and wild-type controls (no. of squares entered in the arena: 9.7 ± 1.3 vs. 9.0 ± 1.6 in males, and 8.7 ± 0.9 vs. 11.1 ± 1.8 in females). About half of the heterozygous mutants (6 of 10 males (χ 2 test: not significant), and 5 of 10 females (χ 2 test: P < 0.05)) and one male control animal showed weak jerks during the observation. In addition, heterozygous mutants showed increased tail elevation than control mice (7 of 10 vs. 0 of 10 in males (χ 2 test: P < 0.01), and 9 of 10 vs. 5 of 10 in females (χ 2 test: not significant)). All other SHIRPA parameters (body position, palpebral closure, lacrimation, defecation and urination during observation, as well as transfer arousal, gait, pelvic elevation, touch escape, positional passivity, trunk curl, limb grasping, pinna reflex, corneal reflex, startle response, contact righting, evidence of biting and vocalization in and above the arena) were without significant alterations.
Measurement of the forelimb grip strength to evaluate muscle performance was done with a grip strength meter system (Bioseb, France), and three trials were undertaken for each mouse and measurement within one minute. The analysis revealed a small reduction of strength in the heterozygous mutants that is most likely due to the lower body weight. Evaluation of motor coordination and balance in three consecutive trials on the rotarod (Bioseb, France) at an accelerating speed from 4 to 40 rpm for 300 sec with 15 min interval between each trial found heterozygous mutants performing even better. This may also be caused by the lower body weight. The usual improvement in the performance of the task over the three trials was observed in the heterozygous mutants and the wild-type males, but not in the wild-type females. In total, significant differences of the neurological screening were quite subtle and might hint towards a slightly higher excitation state in the heterozygous mutants.

Discussion
Two-to-four month-old Umod A227T mutant mice and 7-10 month-old Umod C93F mutant mice were examined using a systemic and standardized, comprehensive phenotypic analysis. The two mouse lines Umod C93F and Umod A227T harbor different mutations within the Umod gene and exhibit a progressive renal phenotype with impairment of the urinary concentrating ability, strongly reduced fractional excretion of uric acid, altered divalent cation metabolism, and progressive morphological kidney alterations which are found in a similar manner in human UAKD. The clinical features of UAKD in humans are heterogeneous also within affected families harbouring the identical UMOD mutation, irrespective of type and site of the mutation that may be due to variations in the genetic background and/or the environment. The two lines Umod C93F and Umod A227T were independently established on the identical C3H inbred genetic background, housed under the same environmental conditions, and analyzed in standardized experiments.
No line-specific additional sequence polymorphisms can be the cause for phenotypic alterations detected in both Umod mutant lines. Thus, they are models not only for determining the mutation-specific impact on the renal dysfunction in UAKD, but also to screen for additional primary and/or secondary effects of Umod mutations that may also appear in affected humans.
Genome-wide association studies in humans for alterations in kidney function [18] and blood pressure regulation [19] found UMOD locus variations including promoter variants leading to increased uromodulin expression.
The point mutation of line Umod C93F leads to the amino acid exchange C93F resulting in the loss of the putative disulfide bond C93-C105 (www.uniprot.org/uniprot/Q91X17). In humans, more than 70 UMOD mutations are known to cause UAKD. Many of them affect the amino acid cysteine. One is C106Y corresponding to murine C105. The severity of the kidney alterations and renal dysfunction was similar in Umod C93F heterozygous mutant mice and Umod A227T homozygous mutant mice. Therefore, adult 2-4 month-old Umod A227T heterozygous mutants and homozygous mutants as well as 7-10 month-old Umod C93F heterozygous mutants were analyzed. Umod C93F homozygous mutants with the strongest kidney alterations were not examined in the German Mouse Clinic. In affected humans, usually heterozygous UMOD mutations occur.
The most obvious morphological phenotype in the mutants of both lines is the reduced body weight that appears due to the highly reduced fat mass. Compared to the respective wild-type controls, the mean body weight reduction was 10% in male heterozygous mutants, 13% in female heterozygous mutants, 15% in male homozygous mutants, and 24% in female homozygous mutants of line Umod A227T at 4 months of age. In line Umod C93F , the body weight reduction was 20-29% in male heterozygous mutants, and 27-35% in female heterozygous mutants at 9 months of age [11,12]. The onset of the reduced body weight of line Umod C93F was revealed at the same time as the onset of the clinical kidney alterations. Thus, at least the onset of the lower body weight may not be the consequence of chronic renal insufficiency, as young mutant mice exhibit only mild symptoms of renal dysfunction [11]. Obvious metabolic alterations were not described in the other published mouse models generated for functional studies of Umod which include two knockout lines [3,4] as well as two transgenic lines expressing mutant Umod. The transgenic line expressing C148W human mutant uromodulin exhibited no clinical renal phenotype even in aged mice [8]. The transgenic line expressing the C147W murine mutant uromodulin showed a clinical phenotype with features of UAKD and renal failure with tubular necrosis at an age of 6 months [9]. The metabolic alterations found in our both mutant mouse lines Umod A227T and Umod C93F might be due to a different expression level of uromodulin (endogenous vs. transgene expression) and/or to the different genetic background (C3HeB/FeJ versus C57BL/6 and FVB/N in the transgenic mice, respectively).
Analysis of energy metabolism in line Umod A227T at the age of 3 months under ad libitum conditions revealed that body mass and body temperature were reduced in mutant mice. Metabolic rate was decreased in mutants as expected for lower body mass but food intake was significantly increased. Metabolic fuel utilization was not different as concluded from the respiratory quotient [12]. These results were essentially confirmed in older Umod C93F heterozygous mutants by two independent analyses. It is not clear if this phenotype is the consequence of one major effect or if multiple minor effects, e.g. an increased urinary glucose excretion per gram body weight being one of them, contribute to the altered energy metabolism. Thus, the metabolic and thermoregulatory properties of both lines have to be clarified in further studies.
In addition, the pathogenic effects of the Umod mutations in the kidney were evaluated by comparative genome-wide transcriptome profiling using 21k cDNA microarrays (data not shown). Kidneys of Umod A227T homozygous mutants and Umod C93F heterozygous mutants (n = 4 males each) were compared to the respective wild-type controls. In line Umod A227T at the age of 17 weeks, 104 significantly regulated genes were identified. The range of the mean log 2 ratios was 0.55 to 1.34 for the 54 up-regulated genes and -1.05 to -0.54 for the 50 down-regulated genes. Several significantly regulated genes are annotated with e.g. hemolytic anemia, glomerulosclerosis, hypertension, diabetes and tumorigenesis. In line Umod C93F at the age of 38 weeks, 54 significantly regulated genes were identified. The range of the mean log 2 ratios was 2.57 for the single up-regulated gene Scd1 (stearoyl-coenzyme A desaturase 1, which also represents the gene with the highest up-regulation in line Umod A227T ) and -2.59 to -1.36 for the 53 down-regulated genes. Several regulated genes are functionally associated with e.g. cholesterol and fatty acid metabolism, proteolysis and apoptosis as well as kidney dysfunctions. These data need to be functionally verified in further experiments. Scd1 is expressed in various tissues and plays a key role in the general energy metabolism [20]. Further studies have to reveal the role of the increased expression of Scd1 in the kidneys of Umod mutant mice.
Due to the altered kidney function and the consecutive obvious alterations of the respective blood plasma parameters found in our two Umod mutant lines, secondary effects in addition to the skeletal and metabolic alterations were suggested to putatively appear especially in the older Umod C93F mutants, like e.g. alterations in hematological, blood gas or blood pressure parameters, chronic inflammation of the gastrointestinal system, or uremic effects on the central nervous system.
Hematological analysis revealed a moderate microcytic and erythropenic anemia in Umod C93F mutants. Blood gas analysis as well as the cardiovascular and lung function data of Umod mutants at the age of 9 months showed no obvious alterations. It remains to be clarified whether the altered plasma ANP values as marker for cardiac filling pressures that appeared to be increased by tendency in Umod mutants of both lines may be also caused by the lower body weight. At least no functional consequences thereof were observed in the cardiovascular analyses. The behavioral (data not shown) and neurological results of both Umod A227T and Umod C93F mutants were within the physiological range of C3H mice and did not permit to define an altered behavioral pattern for mutant animals. The appearance of weak jerks during the observation in about half of the Umod C93F heterozygous mutants remains to be further analyzed.
Immunological analyses of older Umod mutants revealed small changes in the frequencies of leukocyte populations. Furthermore, a higher proportion of CD62L expressing cells within the T cell cluster was found that represent the naive T cell compartment, newly produced in the thymus. This is an often observed phenotype in various mutant lines. In total, no indications were observed for the appearance of major inflammatory processes in the mutant mice. Urinary UMOD excretion was similar in Umod A227T homozygous mutants and Umod C93F heterozygous mutants, and markedly decreased compared to wild-type controls [11]. Recently the role of UMOD as a regulator of renal and systemic immunity was proposed as homozygous Umod knockout mice showed absolutely and relatively enlarged spleens with white-pulp macrophage infiltration as well as an increase of serum cytokines via decreased glomerular clearance and loss of urinary cytokine trapping [21]. The plasma parameters measured in Umod A227T and Umod C93F mutants did not give hints to the proposed role of UMOD as a regulator of immunity in the setting of UAKD with present UMOD expression but disturbed protein maturation. In addition, spleen weights relative to the body weight or carcass weight showed no differences in Umod A227T homozygous mutants and wild-type controls at 3 months and 22 months of age (not shown). Analogously, Umod C93F heterozygous mutants with severely decreased urinary UMOD excretion and wild-type controls at 1 year of age and 84-99 weeks of age showed the same spleen weight relative to the carcass weight (1 year of age (mean ± SD): 0.73 ± 0.05% vs. 0.80 ± 0.15% in males, and 1.08 ± 0.18% vs. 1.04 ± 0.13% in females; n = 8-11 per genotype and sex).
In conclusion, uromodulin is selectively expressed in the kidney and amino acid-changing mutations of Umod lead to the clinical symptoms of UAKD, a slowly progressive renal disease of mild urinary concentration defect, reduced fractional excretion of uric acid and morphological tubulointerstitial kidney alterations. The systemic phenotypic analysis of the two mutant mouse lines Umod A227T and Umod C93F on the C3H inbred genetic background reproducibly revealed alterations in body weight, body composition, bone metabolism, and energy metabolism. In older mutants, a moderate microcytic and erythropenic anemia was revealed. The other analyses in 7-10 months-old mice showed single small additional effects. Thus, depending on the genetic background, long-term effects additional to the kidney lesions might also appear in affected humans. Table S1.

Supporting Information
Time points of the phenotypic analyses described for the lines Umod A227T and Umod C93F in the German Mouse Clinic (GMC). (DOCX)