| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Division of Endocrinology, Department of Medicine, and The Ilyssa Center for Molecular and Cellular Endocrinology, The Johns Hopkins University School of Medicine, Baltimore Maryland 21287
Address all correspondence and requests for reprints to: Roberto Salvatori, M.D., Division of Endocrinology, Johns Hopkins University School of Medicine, 1830 East Monument Street #333, Baltimore, Maryland 21287. E-mail: salvator{at}jhmi.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The majority of children with IGHD grow when treated with GHRH therapy, showing normal function of the GHRH-R and of the somatotroph cells (10). Therefore, GHRH would be a likely candidate gene as a cause of inherited IGHD. However, despite extensive searches, mutations in the GHRH gene have not been identified, and this gene has been excluded by linkage analysis and direct gene analysis in 23 unrelated IGHD families (11). One possible explanation is that the lack of GHRH has no ill effect because other factors substitute for it. Conversely, it is possible that GHRH has other functions, in addition to controlling GH secretion, and that lack of GHRH may cause a broader phenotype than the one observed in GHRH-R mutations or be lethal for the embryo. Because the lack of functional GHRH-R causes IGHD, this hypothesis would imply that GHRH would bind to a different receptor in extrapituitary sites. The possibility of the existence of non-GH-related functions of GHRH is supported by the finding that, in humans, immunoreactive GHRH is present in several regions of the brain (hypothalamus, septum, and substantia innominata), in the mucosal epithelium of the gastrointestinal tract, and in the pancreas (12, 13). In rodents, using RT-PCR, GHRH transcripts have been detected in the cortex, in the brain stem, and in the placenta and testis (14, 15). Using RT-PCR associated with Southern blot hybridization, GHRH transcripts have been detected in several other tissues, such as duodenum and kidney. Given the large size of these organs in comparison to the few hundred GHRH-secreting neurons in the ventromedial and arcuate nuclei of the hypothalamus, it is likely that the total amount of GHRH produced in the hypothalamus represents only a small fraction of the amount produced by the whole body. Accordingly, in humans, GHRH circulates in picograms per milliliter concentration, and levels are similar in normal subjects compared with patients with GH-secreting tumors (16) in whom chronically elevated levels of serum GH and IGF-I are expected to suppress hypothalamic GHRH production by negative feedback (17, 18).
To investigate the functions of GHRH and/or to create a mouse model of IGHD due to lack of hypothalamic GHRH, we have created a mouse carrying a targeted disruption [knockout (KO)] of the GHRH gene (GHRHKO).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To abolish expression of biologically active mature GHRH, we have chosen a classical gene ablation strategy. We have created a plasmid construct containing a mouse GHRH gene in which the neomycin resistance cassette (Neor) was used to replace part of intron 2 and most of exon 3 (Fig. 1A
). We used high-fidelity PCR (Takara LA PCR Kit 2.1; Panvera, Madison, WI) to amplify an approximately 8-kb fragment (spanning from the –820 from the transcription initiation site to the end of exon 4) of mouse 129SV genomic DNA following the manufacturer’s recommendations. This fragment was subcloned into pcDNA 1.0 Amp plasmid. We then screened the 8-kb area with several restriction enzymes and confirmed (as per the Ensembl web site published sequence) that MluI does not cut anywhere in this segment of the gene. Taking advantage of sequences that required only one two-base change per site, two unique MluI sites were introduced via site-directed mutagenesis in the GHRH gene (one in intron 2, 128 bp from the intron 2-exon 3 junction, and one in exon 3, 52 bp distal to the same junction), and the intervening sequence (180 bp) was excised using MluI digestion. The removed fragment corresponded to the last part of intron 2 and about one half of exon 3 and encoded for the last three amino acids of the N-terminal signal peptide and the first 14 amino acids of the mature 1–42 GHRH protein. We reasoned that this strategy would abolish most of GHRH biological activity because the deletion of the initial three amino acid residues from the 1–44 human GHRH [GHRH (3–44)-NH2] reduces hormone activity to 0.016% of full-length GHRH (1–44)-NH2 (22).
|
To add a negative selection marker that would increase the efficiency of screening for homologous recombination, we then inserted the herpes simplex virus thymidine kinase gene at the 3' end of the target vector. In case of homologous recombination, the herpes simplex virus thymidine kinase cassette would be excised, thereby making the cells resistant to the toxic effects of ganciclovir, which requires thymidine kinase.
Creation of the KO mouse
The DNA plasmid containing the recombinant GHRH gene was linearized and then transfected into 129SV embryonic stem (ES) cells by electroporation. Two hundred thirty clones that survived dual antibiotic (neomycin and ganciclovir) selection were screened via long-range PCR (Takara LA PCR Kit 2.1) with a sense primer annealing to the Neor gene (C in Fig. 1A
: 5'-CTTGCCGAATATCATGGTGGAAAATGGCCGCT-3') and an antisense primer annealing to intron 4 sequence located at the 3' end of the vector’s insertion site (D in Fig. 1A: 5'-CCCCATGGGACTACTTACAGATGGTCGTGAG-3'). This reaction is predicted to generate a 3.7-kb fragment in case of homologous recombination. One of the clones exhibited the expected fragment (Fig. 1B, left). DNA from this clone was then analyzed by PCR using a sense primer that anneals to GHRH gene sequences (A in Fig. 1A) that are outside of the 5' end of the site of integration (5'-ACACACCGGAAGCTGGGCATGGTGGCACATG-3') and an antisense primer that anneals to the Neor gene (B in Fig. 1A: 5'-TGCTCGACATTGGGTGGAAACATTCCAGGCC-3'). As one would expect in case of homologous recombination, this reaction generated a 4.8-kb DNA fragment (Fig. 1B, right).
To confirm the PCR experiments and to verify that only one copy of the gene was incorporated in the genome, Southern blot analysis was used. We knew from the Ensembl mouse genome sequence that no SphI restriction site is present in the fragment of GHRH gene corresponding to the targeting vector, but two sites are present immediately outside the area covered by the construct, 8.5 kb from each other. Using SphI-digested normal mouse DNA and a GHRH probe (A in Fig. 1A), we found, by Southern blot analysis, a band of approximately 8.5 kb in length (data not shown). This fits with the predicted location of the SphI sites. In case of homologous recombination, this large DNA sequence would be cleaved into two fragments because of the presence of an additional SphI site within the Neor gene (at position 823 of 1169 bp). Thus, Southern blot analysis using a Neor probe (Fig. 1, probe B) would show two bands, the sum of which would equal 9.5 kb (8.5 kb of endogenous gene plus 1.2 kb of Neor minus 0.2 kb of deleted native gene), one band of 5.8 kb and one band of 3.7 kb. This was confirmed experimentally, as shown in Fig. 1C.
The blot was then stripped and sequentially reprobed with two GHRH probes (Fig. 1A, probe A and probe C). Using this approach, we have obtained bands of the same size as the ones generated by the Neor probe, and therefore, we have confirmed homologous recombination (data not shown). The positive colony was confirmed to contain normal karyotype (euploid with 40 chromosomes).
The positive GHRHKO ES clone was then injected into C57BL6 blastocysts, and the blastocysts were transferred to CBA/F1 foster mothers. A significant percentage of the blastocysts developed into chimeric pups. The high percentage male chimeras were then mated with C57BL6 mice to produce germline transmission animals (male chimeras are more likely to transmit the transgene than female chimeras). Sixty-nine mice were born of agouti color. After weaning, their tails were screened for genotyping. We could not repeat long-range PCR (even of the normal allele) from tail DNA (despite extracting DNA with three different methods), possibly due to the presence of inhibitors of the amplifying enzyme in the tail extracts. Therefore, tails were screened by Southern blotting. Because in our hands SphI (from two different manufacturers: New England Biolabs, Beverly, MA; and Life Technologies, Rockville, MD) did not cut even the normal allele in DNA extracted from the tail, we used a different strategy to screen mice for the presence or absence of the transgene. Because homologous recombination had already been proven in the ES cell, we used a BamHI cut. The Neor gene contains a BamHI site. Therefore, the KO allele, when probed with a probe located 3' of the Neor, will generate a 1.7-kb band, whereas the normal allele will generate a 5.1-kb band. Twenty-five mice were confirmed to be heterozygous for the GHRH KO allele. The disruption of a single GHRH allele did not have any easily identifiable effect on the mice phenotype.
Generation of homozygous KO animals
Offspring were generated by mating heterozygous (+/–) males and females. Fertility and litter size were normal. By d 12, offspring were ear tagged, and tails were cut for genotyping. Pups were weaned at 3 wk of age and housed by sex but not by genotype.
Animals were weighted once a week starting from their second week of life. Body weights were recorded to the nearest 0.1 g using a daily-calibrated electronic balance (Scout Pro Balance; Ohaus Corp., Pine Brook, NJ). A subset of +/– and –/– mice (four males and four females in each group) was anesthetized at 6 wk of age to obtain accurate length (nose to tail root) measurement. Measurements were performed with an Electronic Digital Caliper (Control Co., Friendswood, TX).
All mice experienced a controlled environment with 14-h light, 10-h dark cycles at 21 C and 23% humidity and with standard mouse/rat food (Prolab RMH2500; PMI Nutrition International, Brentwood, MO) and water ad libitum.
Animals were euthanized with a halothane overdose (Sigma Chemical Co., St. Louis, MO). Blood was obtained either by cardiac puncture in euthanized animals or by tail vein puncture in live animals.
All procedures were approved by the Johns Hopkins Institutional Animal Care Committee.
Southern blot analysis of genomic DNA
Genomic DNA was extracted from ES cells by phenol-chloroform extraction and from tail clips using the DNeasy Tissue Kit (Qiagen Inc., Valencia, CA) following the manufacturer’s recommendations. DNA was then digested with restriction enzyme for 12–14 h, separated on a 0.8% agarose gel, transferred onto a nylon membrane (GenescreenPlus; PerkinElmer Life Sciences, Inc., Boston, MA), and prehybridized for 4 h at 42 C following the manufacturer’s recommendations. Hybridization was carried out at 45 C with 32P-labeled cDNA probes. The autoradiography procedure was enhanced by exposing film at –70 C (Omat AR film; Kodak, Rochester, NY) for 8–36 h.
Pathological exam of the pituitary area
Two adult nulliparous females (one +/+ and one –/–) were killed and decapitated. The heads were then skinned, and parietal bones were removed. Heads were fixed in 10% buffered formalin for 48 h. They were rinsed for 2 h in water and then decalcified for 1 wk in 22.5% formic acid/0.45 M sodium citrate. The heads were then rinsed in water and placed in 70% ethanol for 2 h. The area containing the pituitary gland was isolated and then embedded in paraffin. Midline sagittal sections were then cut and stained with hematoxylin and eosin.
Analysis of GHRH mRNA in KO animals
To confirm that mRNA encoding for GHRH was absent in –/– animals, several tissues (cortex, hypothalamus, duodenum, pancreas, kidney, testis, and liver) were harvested from male –/– and +/+ adult animals (and placentas from +/+ and –/– fetuses) (the latter generated by a couple of –/– parents), and total RNA was isolated using TRIzol Reagent (Invitrogen) according to the manufacturer’s protocol. One adult mouse of each genotype was used to extract RNA from hypothalamus, testis, cerebral cortex, duodenum, liver, pancreas, and kidney. Placentas were extracted from a pregnant –/– female on the 16th postcoital day after mating with a –/– male.
Two micrograms of total RNA were used to generate cDNA using reverse transcriptase. Control tubes were used without reverse transcriptase (reverse transcriptase negative). GHRH cDNA was amplified by PCR using the following primers: sense (annealing to exon 2 sequences): 5'-GCTCTGGGTGCTCTTTGTGATC-3', and antisense (annealing to exon 5 sequences): 5'-GGCTCAAGCGTCCGCTGAAGGCTT-3'. These primers were expected to generate a 312-bp band corresponding to the entire coding region of the pre-pro-GHRH. In a parallel reaction, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA was amplified using primers expected to generate a 210-bp band to ensure quality of the cDNA.
GH mRNA measurement
For each experiment, pituitary glands from 9- to 10-wk-old animals (four +/+ animals, four +/– animals, and four –/– animals for each experiment) of both sexes were harvested and pooled, and total RNA was isolated using TRIzol Reagent. Pituitary glands from +/+ and +/– animals yielded an average of 6.1 and 5.6 µg of total RNA per mouse, respectively, whereas pituitary glands from –/– animals were macroscopically smaller and yielded an average of 0.8 µg of total RNA per mouse.
Pituitary GH mRNA content was quantified by Northern analysis of 3 µg of total RNA using a 671-bp 32P-labeled mouse cDNA probe obtained by PCR amplification of GH sequences from mouse pituitary cDNA (sense: 5'-TCCTGACCGTCAGCCTGCTCT-3'; antisense: 5'-GAGGCACAGGAGAGTGCAGCA-3'). GH mRNA band intensity was quantified by a phosphor imager (Molecular Imager FX; Bio-Rad Laboratories, Hercules CA), and results were normalized by comparison to GAPDH expression, which was measured by hybridization (after stripping of the blots) with a 32P-labeled mouse GAPDH cDNA probe. The experiment was repeated three times (for a total of 12 animals/group) with three separately labeled probes. To control for different band intensity depending of the variable specific activity of probes made on different days, counts of the +/+ mice were considered 100%. Results are expressed as means ± SD of the three separate experiments. In a separate experiment, prolactin mRNA content in pituitary from animals with different genotypes was assayed by Northern blot analysis.
GH and prolactin protein measurement
Pituitaries from 10-wk-old +/+ and –/– males were pooled (four glands/group), and proteins were extracted using TRIzol Reagent according to the manufacturer’s recommendations. Protein concentration was determined using the bicinchoninic acid method (Micro BCA Protein Assay Kit; Pierce, Rockford, IL). Five micrograms of proteins of each sample were boiled in 2x sample buffer and resolved in 15% SDS-PAGE and then electrotransferred onto polyvinylidene difluoride membrane (Immobilon-P; Millipore Corporation, Bedford, MA). Five micrograms of protein from the testis of a +/+ adult mouse were used as negative control. To compare GH and prolactin, each sample was loaded in duplicate on the same gel to maintain identical running and transferring conditions. After overnight blocking in Tris-buffered saline plus 0.02% Tween 20 plus 5% milk, half of the membrane was incubated for 2 h at room temperature (RT) with antimouse GH from rabbit (National Hormone and Peptide Program, Harbor University of California Los Angeles Medical Center, Torrance, CA), and the other half was incubated with antimouse prolactin (from the same source), both at 1:80,000 dilution. After washing, both membranes were incubated for 1 h with horseradish peroxidase-conjugated goat antirabbit IgG (1:3000 dilution; Santa Cruz Biotechnology Inc., Santa Cruz, CA). The membranes were washed, and proteins were detected by enhanced chemiluminescence (Amersham Biosciences, Chicago IL). Band size was compared to the Full Range Rainbow protein weight marker (Amersham Biosciences).
Serum and liver IGF-I
Serum IGF-I levels were measured from 9-wk-old animals (five animals for each group and sex) by Anilytics Inc. (Gaithersburg, MD) using rat IGF-I RIA (DSL-2900; DSL, Webster, TX) after acid ethanol extraction, following the manufacturer’s recommendations.
To detect liver IGF-I mRNA, total RNA was extracted from 9-wk-old male mice (four +/+ mice, four +/– mice, and four –/– mice) using TRIzol Reagent following the manufacturer’s recommendations. Analysis was performed by Northern blotting (5 µg of RNA/lane) using 32P-labeled mouse IGF-I cDNA probe on a single blot (kindly donated by Dr. D. LeRoith, National Institutes of Health, Bethesda, MD). Band intensity was analyzed by a phosphor imager, and results were normalized with GAPDH cDNA.
Detection of GHRH-RP
To detect whether GHRH-RP mRNA was present in KO mice, we used RT-PCR. RNA was extracted from testes of +/+ and –/– mice, and cDNA was generated using reverse transcriptase and random primers. Testes cDNA was amplified by PCR using primers (sense: 5'-CCA-GGA-AGA-CAG-CAT-GTG-GAC-AGA-3'; and antisense: 5'-GGC-TCA-AGC-GTC-CGC-TGA-AGG-CTT-3') designed to amplify the exon 4 and 5 sequences, encoding for GHRH-RP. We found the expected size amplicon both in +/+ and –/– mice and no band in the reverse transcriptase-negative controls (data not shown).
To determine whether the GHRH-RP message was translated appropriately, Western blot analysis was performed. Protein extracts from testes of 9- to 10-wk-old animals were obtained with TRIzol Reagent according to the manufacturer’s recommendation. To compare detection of the peptide with specific antibody and preimmune serum, samples were loaded in duplicate and separated on a discontinuous polyacrylamide tricine-sodium dodecyl sulfate gel system and electrotransferred onto polyvinylidene difluoride membrane (Immobilon-P; Millipore Corporation). Blocking was carried out at RT for 2 h in Tris-buffered saline plus 0.05% Tween 20 (Sigma) plus 3% nonfat dry milk (Bio-Rad). Half of the membrane was incubated for 2 h at RT with a polyclonal primary antibody from rabbit directed against rat GHRH-RP (generously donated by Dr. O. Pescovitz, University of Indiana, Indianapolis, IN), and half of the membrane was incubated with preimmune serum. After washing steps, both membranes were incubated in blocking solution for 2 h at RT with horseradish peroxidase-conjugated goat antirabbit-IgG (Santa Cruz Biotechnology), and peptide was visualized with enhanced chemiluminescence reagent (Amersham Biosciences).
Statistics
Data were analyzed by ANOVA using SPSS statistical package (SPSS Inc., Chicago, IL), with posthoc analysis using Bonferroni’s method. Data were considered statistically significant at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Interestingly, we noted that 100% of the +/+ pups displayed a deep dark brown coat color, whereas all +/– and –/– mice were agouti color. This is due to the fact that the –/– allele was derived from 129SV agouti color mice and that the gene encoding for agouti (which is inherited as dominant trait) is on chromosome 2, only 2.5 mb from the GHRH gene, and therefore, it segregates with the –/– allele.
Adult –/– mice were autopsied by a veterinarian pathologist who found no anatomic abnormalities in any of the internal organs, with the exception of proportionate small body size and a small pituitary gland.
Male –/– animals had normal copulatory behavior and fertility. When mated with +/– females, no differences in terms of litter size (8.3 pups/litter) or Mendelian ratio was observed (+/–, 48%; –/–, 52%). On the contrary, –/– females, although maintaining normal fertility, had a consistent reduction in litter size (average, 4.1 pups/litter). All –/– adult females displayed normal duration of gestation (19–21 d), normal pup retrieval, and normal maternal behavior. Pups of –/– females had 8.5% mortality in the first 24 h. Failure to thrive was observed in those mice born from primiparous mothers. These mice were not included in the analyses described below; the mice in the analyses were all born from +/– parents.
Growth
All mice analyzed for growth were born from +/– mothers. Growth curves are shown in Fig. 2. At 2 wk of age, –/– mice appeared slightly smaller in size compared with their littermates, but the difference did not reach statistical significance. By wk 3, –/– animals were significantly lighter than their +/+ and +/– littermates, and by wk 8, their weight was reduced to 55–60% of the +/+ mice. This difference was maintained through wk 12. No significant difference was observed at any point between +/+ and +/– mice. A picture of three adult mice with the three different genotypes is shown in Fig. 3 (age and sex matched).
|
|
Pituitary pathology
In agreement with the macroscopic finding of pituitary hypoplasia noted in –/– mice at the time of harvesting of pituitary RNA, Fig. 4 shows a midline sagittal section of a normal adult female +/+ mouse (A) and of a –/– age- and sex-matched control (B), demonstrating evident reduction in size of the anterior pituitary.
|
, PCR amplification of the entire coding region of the GHRH from hypothalamic, testicular, and placental cDNA from a +/+ mouse generated the expected size band (312 bp). The band was absent in RNA extracted from –/– mice and –/– placentas. The –/– cDNA from all samples generated an expected size band (210 bp) corresponding to the housekeeping gene GAPDH, which confirmed that cDNA was of good quality in –/– mice as well.
|
, the pituitary GH mRNA concentration of –/– animals was lower when compared with +/+ mice. There was no statistically significant difference between +/+ and +/– mice. Prolactin mRNA was expressed in –/– mice at levels comparable to +/+ and +/– animals (data not shown).
|
, Western blot analysis using extract from pituitaries pooled from four +/+ and four –/– mice showed a clear reduction in the intensity of the band corresponding to GH (
21 kDa), confirming the mRNA data. As expected, no band was observed from testicular extracts. As a control, equal amounts of protein extracts were used to measure prolactin content. The intensity of the prolactin bands was similar in –/– and +/+ animals.
|
, both male and female –/– groups had significantly reduced serum IGF-I levels compared with their sex-matched +/+ and +/– groups. Surprisingly, +/– males had lower serum IGF-I levels when compared with +/+ males, whereas no differences were found between +/+ and +/– female mice.
|
).
|
).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Confirmation of the importance of GHRH for GH synthesis comes from transgenic and naturally occurring mouse models of GH deficiency (GHD). Three transgenic models of GHRH deficiency are available: 1) Szabo et al. (26) have developed a mouse in which targeted overexpression of GH in the arcuate and periventricular nuclei of the hypothalamus causes IGHD due to negative GH feedback on GHRH expression; 2) a similar model was also created in rats, called transgenic growth retarded (27); however, in these two models, the reduction of GHRH expression is likely limited to the hypothalamus, and hypothalamic GHRH mRNA, albeit reduced, is still detectable (26); and finally, 3) Zhu et al. (28) recently reported a KO mouse in which the production of mature GHRH was altered due to lack of proprotein convertase 1/3 (PC1/3). This enzyme is required for cleavage of pre-pro-GHRH into mature GHRH. Mice lacking this enzyme have reduced GH mRNA content in their pituitaries and low serum IGF-I, resulting in longitudinal growth failure. These mice, however, have high perinatal mortality. They also lack additional hormones that require PC1/3 to be processed to its mature form (such as ACTH, insulin, and glucagons, and possibly other hormones), making it difficult to determine which part of the phenotype is due to the lack of each specific hormone.
Two naturally occurring rodent models of IGHD exist. The dwarf rat, which carries a still undiscovered autosomal recessive mutation (29), and the little mouse, which has a recessive missense mutation in the GHRH-R that causes the receptor to be unable to bind the ligand. In addition, an ethylnitrosourea-induced mouse mutant with a semidominant GH gene mutation has been reported recently (30). However, these models have (or are likely to be) proven unresponsive to GHRH and, therefore, cannot be used to study hypothalamic GHRH deficiency (31).
Hence, to date, there is no clean model of complete lack of GHRH. We had hypothesized that it was possible that generalized lack of GHRH might cause a wider phenotype than IGHD. We also reasoned that, if the phenotype was limited to IGHD, this mouse might be a more appropriate model of IGHD that occurs in humans because most IGHD cases are due to lack of hypothalamic GHRH rather than to a problem in the GHRH-R or in the somatotroph cells (10).
In designing the targeting vector, we have removed the sequences encoding for the initial 14 amino acids of the 1–42 GHRH. Because the initial amino acid residues of the mature human GHRH are essential for its biological activity (22) and there is a high degree of interspecies homology in the N terminus of this molecule, we predicted that the lack of these 14 residues would abolish any GHRH activity.
To confirm that the expression of the normal GHRH mRNA was abolished in KO mice, we used RT-PCR to amplify GHRH mRNA from hypothalamic, testicular, and placental extracts (the only tissue that in our hands showed GHRH mRNA expression). Although cDNA from normal mice yielded the expected size band corresponding to the whole coding region of the pre-pro-GHRH (from exon 2 to exon 5), this band was absent in the RNA obtained from –/– mice, confirming that we successfully ablated the expression of the full-length GHRH mRNA.
When we mated +/– founder mice among themselves, the expected Mendelian proportions in the offspring were respected, suggesting that the –/– embryos are as viable as the +/+ and +/– animals. The homozygous –/– mice had normal phenotype at birth and started showing evidence of growth retardation after the second week of life. This pattern reproduces very closely the one observed in the little mouse. Although +/– females seemed to be heavier than their +/+ littermates, this difference was not significant. Up to the 12th week, the –/– animals showed no sexual dimorphism in weight, and females and males had similar weight.
The marked degree of growth retardation of the GHRHKO animals resembles the growth retardation observed in mice with IGHD of other etiologies. To confirm that the growth retardation of the GHRHKO mice is indeed due to lack of GH, we performed two series of experiments.
First, we measured pituitary GH mRNA content and showed that it is significantly reduced in –/– mice compared with +/+ and +/– animals. The degree of reduction (20% of controls) was comparable to what was observed in PC1/3 KO mice (28). The fact that GH mRNA, albeit low, was still detectable fits with what was observed in the little model, in which somatotroph cells, albeit reduced in number, were present and able to secrete GH (32). Western blot analysis confirmed the marked reduction of GH content in the GHRHKO pituitary, whereas prolactin content was similar to normal animals. Moreover, pathological analysis showed clear evidence of pituitary hypoplasia, which resembles the finding in the little model. Indeed, the average amount of total RNA obtained from each –/– pituitary gland was only 15% of the amount obtained from +/+ and +/– animals.
Second, we measured serum IGF-I. As expected, –/– mice of both sexes had significantly lower serum IGF-I than +/+ animals. Whereas, in females, there was no difference in serum IGF-I between +/+ and +/– animals, the heterozygous male mice had a significant reduction in serum IGF-I compared with +/+ animals, suggesting an effect of GHRH haploinsufficiency. The cause and the biological meaning of this difference are not clear because no difference was observed in liver IGF-I mRNA content, and liver is the source of most of the circulating IGF-I (33). In addition, +/– male mice were not smaller than their +/+ littermates, showing no consequence of this mild reduction in serum IGF-I on growth. This fits with the observation that circulating IGF-I has no direct correlation to somatic growth; mice with liver-specific deletion of the IGF-I gene grow well despite very low serum IGF-I, suggesting that IGF-I produced locally in the growth plates is more important for growth than the circulating molecule (34). It is unclear why the difference in serum IGF-I between +/– and +/+ animals is limited to males. There have been reports of sexual dimorphism in the role of GHRH in control of GH secretion both in rodents and humans (35, 36), and this sex difference may reflect the need for a higher GHRH level to maintain GH secretion in males than in female mice. Finally, such reduction in serum IGF-I was not observed in the +/lit mice (Donahue, L., Jackson Laboratories, Bar Harbor, ME, personal communication). It is possible that, in male mice, haploinsufficiency for GHRH (but not for its receptor) has a biological consequence because, in the normal animals, there is an overabundance of the receptor but not of the ligand.
The phenotype of the GHRHKO mice is similar to the one observed in the lit/lit animals, seemingly indicating that the main function of GHRH is to stimulate GH secretion and that its function is mostly mediated by the receptor expressed in the pituitary. In addition to growth retardation, the lit/lit mouse has other phenotypic characteristics, such as increased fat deposition in late adulthood and microcephaly (37, 38). We have not yet evaluated our mice for these more subtle features.
One question is whether the C terminus GHRH-RP is expressed in our model. The function (if any) of this peptide is unknown, although mice overexpressing it have mild abnormalities in the white cell lineage (39). The region encoding for GHRH-RP was not removed by our gene ablation approach. Because GHRH-RP is highly expressed by the Sertoli cells of the testis, we used RT-PCR and found that GHRH-RP message was present in the testis (data not shown). To confirm that mRNA was in frame and appropriately transcribed, we used Western blot analysis. Testes from –/– mice express GHRH-RP, showing that our model selectively obliterated the function of the mature 1–42 GHRH and leaving the question open of possible functions of the C-terminal peptide. Because exon 3 consists of 102 bp (multiple of 3), it is possible that the lack of the intron splice 2 acceptor (removed in our construct) causes splicing of exon 3 sequences, maintaining in-frame exon 4 and 5 mRNA sequences.
This mouse model demonstrates that generalized ablation of the gene encoding for the mature 1–42 GHRH causes growth retardation due to GHD. This model will be useful to study the effects of different compounds on GH secretion in animals with intact somatotroph cells and intact GHRH-R. The main limit of the use of our model for this purpose is the pituitary hypoplasia and the consequent limited number of somatotroph cells able to respond to such stimuli. We are presently planning to treat the –/– animals with GHRH analogs for several weeks to determine whether we can revert the pituitary hypoplasia and create a model of hypothalamic GHD with normal pituitary mass. Now that we know that generalized lack of GHRH does not seem to affect any other major bodily function, an alternative strategy would be to create a temporal conditional KO model (40), in which GHRH ablation is turned on at different stages of postnatal life.
In conclusion, targeted ablation of the GHRH gene in mice causes IGHD with a phenotype that closely resembles the one observed in other rodent models of GHD. This finding should trigger new searches for mutations in the GHRH gene in children with familial IGHD.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: ES, Embryonic stem; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GHD, GH deficiency; GHRH-R, GHRH receptor; GHRHKO, GHRH knockout; GHRH-RP, GHRH-related peptide; IGHD, isolated GH deficiency; KO, knockout; Neor, neomycin resistance cassette; PC1/3, proprotein convertase 1/3; RT, room temperature.
Received January 30, 2004.
Accepted for publication May 10, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Sagazio, X. Xiao, Z. Wang, M. Martari, and R. Salvatori A single injection of double-stranded adeno-associated viral vector expressing GH normalizes growth in GH-deficient mice J. Endocrinol., January 1, 2008; 196(1): 79 - 88. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Zhu, A. S. Gleiberman, and M. G. Rosenfeld Molecular Physiology of Pituitary Development: Signaling and Transcriptional Networks Physiol Rev, July 1, 2007; 87(3): 933 - 963. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Wang, J. Li, C. Y. Wang, A. H. Y. Kwok, and F. C. Leung Identification of the Endogenous Ligands for Chicken Growth Hormone-Releasing Hormone (GHRH) Receptor: Evidence for a Separate Gene Encoding GHRH in Submammalian Vertebrates Endocrinology, May 1, 2007; 148(5): 2405 - 2416. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Baccam, G. Alonso, T. Costecalde, P. Fontanaud, F. Molino, I. C. A. F. Robinson, P. Mollard, and P.-F. Mery Dual-Level Afferent Control of Growth Hormone-Releasing Hormone (GHRH) Neurons in GHRH-Green Fluorescent Protein Transgenic Mice J. Neurosci., February 14, 2007; 27(7): 1631 - 1641. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Alba, D. Fintini, A. Sagazio, B. Lawrence, J.-P. Castaigne, L. A. Frohman, and R. Salvatori Once-daily administration of CJC-1295, a long-acting growth hormone-releasing hormone (GHRH) analog, normalizes growth in the GHRH knockout mouse Am J Physiol Endocrinol Metab, December 1, 2006; 291(6): E1290 - E1294. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. D. Ward, B. M. Stone, L. T. Raetzman, and S. A. Camper Cell Proliferation and Vascularization in Mouse Models of Pituitary Hormone Deficiency Mol. Endocrinol., June 1, 2006; 20(6): 1378 - 1390. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Segal-Lieberman, H. Rubinfeld, M. Glick, N. Kronfeld-Schor, and I. Shimon Melanin-concentrating hormone stimulates human growth hormone secretion: a novel effect of MCH on the hypothalamic-pituitary axis Am J Physiol Endocrinol Metab, May 1, 2006; 290(5): E982 - E988. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Fintini, M. Alba, and R. Salvatori Influence of Estrogen Administration on the Growth Response to Growth Hormone (GH) in GH-Deficient Mice Experimental Biology and Medicine, November 1, 2005; 230(10): 715 - 720. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Huerta-Ocampo, H. C. Christian, N. M. Thompson, M. M. El-Kasti, and T. Wells The Intermediate Lactotroph: A Morphologically Distinct, Ghrelin-Responsive Pituitary Cell in the Dwarf (dw/dw) Rat Endocrinology, November 1, 2005; 146(11): 5012 - 5023. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Alba, D. Fintini, C. Y. Bowers, A. F. Parlow, and R. Salvatori Effects of long-term treatment with growth hormone-releasing peptide-2 in the GHRH knockout mouse Am J Physiol Endocrinol Metab, November 1, 2005; 289(5): E762 - E767. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. R. Le Tissier, D. F. Carmignac, S. Lilley, A. K. Sesay, C. J. Phelps, P. Houston, K. Mathers, C. Magoulas, D. Ogden, and I. C. A. F. Robinson Hypothalamic Growth Hormone-Releasing Hormone (GHRH) Deficiency: Targeted Ablation of GHRH Neurons in Mice Using a Viral Ion Channel Transgene Mol. Endocrinol., |
