Molecular and Cellular Neuroscience 14, 99 -120 (1999)
`Article ID mcne.1999.0767, available online at http: / /www.idealibrary.com on ID E
`
`E
`
`MCN
`
`Mice Lacking oz-Calcitonin Gene -Related Peptide
`Exhibit Normal Cardiovascular Regulation and
`Neuromuscular Development
`
`Jonathan T. Lu,* Young -Jin Sont Jongho Lee,*
`Thomas L. Jetton,t Masakazu Shiota,t Lisa Moscoso,t
`Kevin D. Niswender,t Arthur D. Loewy,t Mark A. Magnuson,t
`Joshua R. Sanes,t and Ronald B. Emeson *,t
`*Department of Pharmacology and 'Department of Molecular Physiology and Biophysics,
`Vanderbilt University School of Medicine, Nashville, Tennessee 37232: and 'Department of
`Anatomy and Neurobiology, Washington University Medical School, St. Louis, Missouri 63110
`
`a- Calcitonin gene -related peptide («CGRP) is a pleiotropic
`peptide neuromodulator that is widely expressed through-
`out the central and peripheral nervous systems. CGRP has
`been implicated in a variety of physiological processes
`including peripheral vasodilation, cardiac acceleration,
`nicotinic acetylcholine receptor (AChR) synthesis and
`function, testicular descent, nociception, carbohydrate
`metabolism, gastrointestinal motility, neurogenic inflam-
`mation, and gastric acid secretion. To provide a better
`understanding of the physiological role(s) mediated by
`this peptide neurotransmitter, we have generated «CGRP -
`null mice by targeted modification in embryonic stem
`cells. Mice lacking «CGRP expression demonstrate no
`obvious phenotypic differences from their wild -type litter -
`mates. Detailed analysis of systemic cardiovascular func-
`tion revealed no differences between control and mutant
`mice regarding heart rate and blood pressure under basal
`or exercise -induced conditions and subsequent to pharma-
`cological manipulation. Characterization of neuromuscu-
`lar junction morphology including nicotinic receptor local-
`ization, terminal sprouting in response to denervation,
`developmental regulation of AChR subunit expression,
`and synapse elimination also revealed no differences in
`«CGRP- deficient animals. These results suggest that
`«CGRP is not required for the systemic regulation of
`cardiovascular hemodynamics or development of the neu-
`romuscular junction.
`
`INTRODUCTION
`
`Calcitonin gene -related peptide (CGRP) is a 37- amino-
`acid neuropeptide produced via tissue -specific alterna-
`tive splicing of the calcitonin /aCGRP primary RNA
`
`transcript (Amara et al., 1982, 1984; Rosenfeld et al., 1983,
`1984). While calcitonin mRNA production is largely
`limited to the C -cells of the thyroid gland, CGRP
`transcripts are widely expressed in discrete cell types
`throughout the central and peripheral nervous systems
`(Amara et al., 1982; Rosenfeld et al., 1983). In rats and
`human beings, posttranslational processing of calcito-
`nin and CGRP preprohormones is predicted to result in
`the generation of six distinct peptides including calcito-
`nin and amino- and carboxyl -terminal peptides derived
`from the calcitonin prohormone. The aCGRP prohor-
`mone is posttranslationally modified to yield «CGRP
`and N- and C- terminal peptides and biological activities
`have been ascribed to all but the carboxyl -terminal peptide
`of pro -CGRP (Burns et al., 1992). A second isoform of
`CGRP, referred to as 13-CGRP or CGRP II, is encoded by
`a separate gene locus and has been shown to demon-
`strate an overlapping but nonidentical pattern of expres-
`sion throughout the nervous system (Amara et al., 1985).
`Based upon its anatomical distribution and bioactivi-
`ties, CGRP is thought to play important roles in auto-
`nomic, somatosensory, integrative, and motor functions
`(Kawai et al., 1985; Kresse et al., 1992, 1995; Rosenfeld et
`al., 1983; Skofitsch and Jacobowitz, 1985a,b). Several
`lines of evidence suggest that CGRP may participate in
`the regulation of cardiovascular hemodynamics. CGRP
`immunoreactivity has been localized in almost all major
`organs involved in cardiovascular regulation, including
`peripheral perivascular nerves, epicardial arteries, and
`sinoatrial and atrioventricular nodes, as well as CNS
`regions vital for the regulation of cardiovascular horneo-
`
`1044-7431/99 $30.00
`Copyright
`1999 by Academic Press
`All rights of reproduction in any form reserved.
`
`Lilly Exhibit 1288
`Eli Lilly & Co. v. Teva
`Pharms. Int'l GMBH
`
`99
`
`

`

`100
`
`Lu et al.
`
`stasis (nucleus tractus solitarii, vagus nerve, lateral
`parabrachial nucleus, nucleus ambiguus, central amyg-
`daloid nucleus, and hypothalamus) (Hokfelt et al., 1992;
`Mulderry et al., 1985; Shoji et al., 1987; Terrado et al.,
`1997; Zaidi et al., 1985). High -affinity CGRP receptor
`sites have been demonstrated in the media and the
`intima of resistance vessels as well as in the heart muscle
`itself, including the nodal and conductive system and
`the pacemaker cells (Coupe et al., 1990; Franco -Cereceda
`et al., 1987b; Ono et al., 1989; Sigrist et al., 1986) . Systemic
`administration of synthetic CGRP is found to be pro-
`foundly hypotensive in multiple species (Beglinger et
`al., 1991; Ezra et al., 1987; Franco -Cereceda et al., 1987a;
`Okamoto et al., 1992; Young et al., 1993), leading to the
`speculation that CGRP may be an important regulator of
`systemic vascular tone (McEvan et al., 1989). Paradoxi-
`cally, intracerebroventricular injection or microinjection
`of CGRP into specific regions of the CNS elicited
`significant increases in mean arterial pressure (MAP)
`and tachycardia, presumably due to an increase in
`sympathetic outflow (Fisher et al., 1983; Hasegawa et al.,
`1993; Kuo et al., 1994; Nguyen et al., 1986).
`CGRP has also been implicated in the development
`and function of the skeletal neuromuscular junction.
`is synthesized in vivo by developing motor
`CGRP
`neurons, axonally transported to nerve terminals, stored
`in large dense core vesicles (Kashihara et al., 1989;
`Matteoli et al., 1988), and released upon nerve stimula-
`tion (Sakaguchi et al., 1991; Sala et al., 1995). Levels of
`CGRP decline in adulthood, but are markedly increased
`following axotomy (Dumoulin et al., 1992; Moore, 1989)
`or paralysis (Sala et al., 1995), situations in which
`synapse formation is favored. Exogenously applied
`CGRP can act presynaptically to increase quantal size
`and output at the frog neuromuscular junction (Kloot et
`al., 1998), while application to cultured myotubes stimu-
`lates accumulation of acetylcholine receptors (AChRs)
`(Fontaine et al., 1986; Moss et al., 1991; New and Mudge,
`1986) by increasing transcription of AChR subunit genes
`(Fontaine et al., 1987; Osterlund et aI.,1989) . Accordingly,
`it has been proposed that CGRP is a nerve -derived
`activator of AChR transcription (Duclert and Changeux,
`1995). Moreover, CGRP increases the agonist- induced
`desensitization rate (Mulle et al., 1988) and channel open
`time (Lu et al., 1993) of AChRs in cultured myotubes,
`suggesting that it affects AChR- mediated signaling at
`developing and regenerating junctions. Finally, applica-
`tion of CGRP to paralyzed muscles suppresses the nerve
`sprouting that would otherwise occur (Tsujimoto and
`Kuno, 1988), suggesting that CGRP is involved in the
`regulation of axonal growth during reinnervation (Sala
`et al., 1995).
`
`The observation that CGRP can elicit numerous re-
`sponses in vitro and in vivo has led to multiple hypoth-
`eses regarding the role of this peptide in diverse physi-
`ological systems. The lack of specific CGRP antagonists
`(Dumont et al., 1997), the characterization of multiple
`CGRP receptor subtypes (Aiyar et al., 1996; Egerton et
`al., 1995; Kapas and Clark, 1995; Luebke et al., 1996;
`Poyner, 1997), and the overlapping expression pattern
`for the a- and f3-CGRP peptides has further complicated
`studies regarding the biological relevance of these neu-
`rotransmitter molecules. To examine the physiological
`role(s) of «CGRP, we have generated and characterized
`mutant mice in which the ability to express «CGRP has
`been selectively ablated. In the present study, we have
`focused upon the proposed role of «CGRP as a regulator
`of systemic cardiovascular hemodynamics and as a
`nerve -derived modulator of neuromuscular develop-
`ment.
`
`RESULTS
`
`Generation of «CGRP -Deficient Mice
`While many studies of mouse physiology using gene
`targeting strategies have relied upon insertion of a
`neomycin resistance cassette to disrupt the coding
`potential of a specific genomic locus, the fact that the
`calcitonin /aCGRP gene encodes two distinct prohor-
`mones via alternative RNA processing raises concerns
`regarding the use of a traditional "knockout" strategy
`for the production of «CGRP -null mice. Knockout of the
`entire gene would result not only in the loss of «CGRP
`expression, but also in the loss of calcitonin as well as
`the amino- and carboxyl -terminal peptides derived
`from each prohormone, making interpretations of any
`resultant phenotype problematic. To circumvent these
`difficulties, a strategy was employed to specifically
`eliminate CGRP expression, while leaving production of
`the remaining five calcitonin /«CGRP- derived peptides
`intact (Fig. 1A). In this strategy, a stop codon (TGA) was
`introduced by oligonucleotide -directed site mutagen-
`esis (Ausubel et al., 1989) before the CGRP- encoding
`region within exon 5; this modification alone results in
`the loss of both CGRP expression as well as the 4- amino-
`acid peptide (DLQA) derived from the carboxyl -
`terminus of the aCGRP prohormone. Although this
`carboxyl -terminal peptide has no known biological
`function, the targeting vector was further modified to
`include an additional copy of this peptide immediately
`following the proteolytic processing signal (Lys -Arg)
`normally found between the N- terminal (N- proCGRP)
`and the CGRP peptides (Fig. 1B). This additional modi-
`
`

`

`Cardiovascular and NMJ Analysis of aCGRP-Null Mice
`
`101
`
`A
`
`H
`
`E C
`
`El
`X
`
`KX r
`
`X
`
`A
`
`B
`WT gene -1-1A
`
`Targeting Vector
`
`Modified gene --\
`
`B
`
`E B H 1rI
`A x
`
`K
`
`E
`
`jX
`(
`
`'
`
`X X
`
`KX E B H
`
`3
`
`1
`
`1
`
`X X
`
`A
`
`B
`
`V
`GTC
`GTC
`V
`
`A
`T
`ACT
`GCT
`GCT
`ACT
`A
`T
`
`Q
`CAG
`CAG
`Q
`
`'K
`AAG
`AAG
`K
`
`R'
`AGA
`AGA
`R
`
`S
`TCC
`GAC
`D
`
`C
`TGC
`CTT
`L
`
`N
`AAC
`CAG
`GCC
`Q A
`
`TGA ggt tac cac
`
`wild -type
`CGRP -null
`
`T
`ACT
`act
`
`A
`GCC
`gcc
`
`T
`ACC
`acc
`
`C
`TGT
`tgt
`
`V
`GTG
`gtg
`
`T
`ACC
`acc
`
`H
`CAT
`cat
`
`R
`CGG
`cgg
`
`L
`CTG
`ctg
`
`A
`GCA
`gca
`
`G
`S
`L
`L
`GGT CTG CTG AGC
`ggt ctg ctg agc
`
`R
`AGA
`aga
`
`G
`S
`GGA
`TCA
`tca
`gga
`
`G
`GGT
`ggt
`
`V
`GTG
`gtg
`
`V
`GTG
`gtg
`
`K
`AAG
`aag
`
`D
`GAC
`gac
`
`N
`AAC
`aac
`
`F
`TTT
`ttt
`
`V P
`N
`T
`GTT CCC ACC AAT
`gtt ccc ace aat
`
`G
`GGC
`ggc
`
`S
`TCT
`tct
`
`A
`E
`GAA GCC
`gaa gcc
`
`F
`TTC
`ttc
`
`G
`GGC
`ggc
`
`'R
`CGC
`cgc
`
`R
`CGC
`cgc
`
`R
`CGC
`cgc
`
`R 101111111121
`AGG GAC CTT CAG
`agg gac ctt cag
`
`*
`TGA
`tga
`
`V
`GTG
`gtg
`
`0 GCC
`
`gcc
`
`C
`
`1r+rrimp
`Irr
`
`010 el
`+/+ +/- +/+ +/- -/-
`
`aft
`
`4-- wild -type allele (205 bp)
`
`CGRP -null allele (185 bp)
`
`-/- +/- : genotype
`
`FIG. 1. Targeting strategy for the generation of aCGRP -null mice. (A) Schematic diagram and abbreviated restriction map of the mouse
`calcitonin /aCGRP gene before and after targeted gene modification. The location of exon 5 is indicated. A, Apal; B, BstEII; E, EcoRI; H, HindIII; K,
`Kpnl; X, Xbal. (B) Nucleotide and amino acid sequence comparison between the wild -type and the aCGRP -null alleles within the coding region of
`exon 5. Introduction of a stop codon (Y) immediately upstream of the CGRP coding region and duplication of the predicted 4 -amino -acid
`carboxyl -terminal peptide (DLQA) after a dibasic proteolytic cleavage site (KR) result in the selective loss of CGRP peptide expression.
`Introduction of a BstEII restriction site (ggttacc), for genomic screening purposes, is double -underlined and 3 -untranslated information is
`presented in lowercase letters. (C) Analysis of mouse genotype by polymerase chain reaction amplification of mouse tail genornic DNA and
`subsequent digestion with BstEII. The migration positions for the wild -type ( + / +; 205 bp) and aCGRP -null ( -/ -; 185 bp) alleles are indicated as
`is the determined mouse genotype.
`
`fication results in the production of an RNA species
`encoding a normal calcitonin prohormone and a modi-
`fied CGRP prohormone in which N- proCGRP and the
`carboxyl- terminal peptide can be produced as a result of
`posttranslational proteolytic cleavage. A BstE II restric-
`tion site, downstream from the engineered stop codon,
`was introduced to alter the translation frame as well as
`facilitate analyses of homologous recombination in tar-
`geted embryonic stem cells and resultant mouse strains
`(Figs. lA and 1B).
`The aCGRP -null mutation was introduced into the
`embryonic stem cells via a replacement -type targeting
`vector (Hogan, 1994) (Fig. 1A). The integration of the
`targeting vector and the fidelity of homologous recombi-
`
`nation were confirmed by Southern blot analyses using
`5 -, 3 -, internal, and neo- specific probes and the nucleo-
`tide sequence surrounding the targeted mutation was
`verified by DNA sequence analyses of PCR- amplified
`ES cell genomic DNA (data not shown). Targeted ES cell
`lines were introduced into C57BL /6 blastocysts and the
`blastocysts were transferred to the uterus of a pseudo -
`pregnant foster mother (ICR; The Jackson Laboratories)
`and allowed to develop normally. Five male offspring
`were produced demonstrating an agouti coat color
`chimerism ranging from 40 to 95 %. These males were
`mated with Black Swiss females and two of the five
`founder animals demonstrated germ -line transmission
`of the agouti coat color and the modified aCGRP allele.
`
`

`

`102
`
`Lu et al.
`
`The genotype of mice (129Sv x Black Swiss) bearing the
`aCGRP-null mutation was determined by restriction
`endonuclease digestion of PCR amplified genomic DNA
`and animals heterozygous for the aCGRP-null allele
`were mated to generate animals homozygous for the
`modified allele (Fig. 1C). Mice that were wild -type,
`heterozygous, and homozygous for the «CGRP -null
`mutation were generated with a Mendelian distribu-
`tion. Homozygous aCGRP -null animals were found to
`be viable, fertile, and outwardly indistinguishable from
`their wild -type littermates.
`
`Molecular and lmmunohistochemical
`Characterization of aCGRP -Null Mice
`The molecular consequences of the aCGRP-null muta-
`tion on CGRP and calcitonin expression were investi-
`gated at both the RNA and the peptide levels using
`ribonuclease protection and immunohistochemical
`analysis. RNase protection studies of total RNA isolated
`from whole mouse brain using a probe designed for the
`wild -type calcitonin/aCGRP allele (Fig. 2A, left), indi-
`cated that control mice expressed only wild -type CGRP
`mRNA transcripts, heterozygous animals expressed both
`wild -type and modified CGRP mRNAs, and aCGRP-
`
`null homozygotes expressed only modified CGRP tran-
`scripts (Fig. 2B, left). A second RNase protection strat-
`egy, using an antisense probe corresponding to the
`modified «CGRP allele (Fig. 2A, right), provided results
`in agreement with those performed using the wild -type
`probe (Fig. 2B, right) . Phosphorlmager analyses of
`protected bands in heterozygous animals indicated that
`the steady -state expression level for the modified CGRP
`mRNA was only 20 -25% of that observed for the
`wild -type transcript (Figs. 2A and 2B), suggesting that
`the modified RNA could be less stable than its wild -type
`counterpart. RNase protection analysis of total RNA
`isolated from thyroid gland showed that both wild -type
`and homozygous mutant animals expressed predomi-
`nantly calcitonin mRNA (Fig. 2C), indicating that the
`aCGRP -null mutation did not perturb the tissue -specific
`alternative splicing of calcitonin /«CGRP primary RNA
`transcripts nor did it significantly alter the level of
`calcitonin mRNA expression compared with a [3-actin
`internal control.
`To confirm that «CGRP peptide expression was ab-
`lated in mutant animals, immunohistochemical analy-
`ses were performed using an amide -specific CGRP
`antiserum that recognized both a- and [3-CGRP (Rosen-
`blatt and Dickerson, 1997). Robust CGRP -like immuno-
`
`A
`
`wild -type RNase probe:
`Exon 5
`
`Exon 4
`
`czCGRP-null RNase probe:
`
`Exon 4
`
`Exon 5
`
`1
`
`441
`
`1
`
`1
`
`452
`
`1
`
`165
`
`calcitonin
`
`CGRP
`
`uCGRP-null
`
`I
`
`184
`
`158
`
`I
`
`165
`
`1
`
`calcitonin
`
`CGRP
`
`aCGRP-null
`
`I 158
`
`195
`
`«CGRP -null
`
`1I
`-/- +/+ +/-
`
`I
`
`-/-
`
`B
`
`wild -type
`
`I
`
`+1+ +/-
`
`uCGRP
`
`aCGRP-null
`
`: probe C
`: genotype : +/+ -/-
`
`uCGRP-null
`
`cxCGRP
`
`calcitonin
`
`p -actin
`
`FIG. 2. Ribonuclease protection analysis of «CGRP and calcitonin mRNA expression in wild -type and aCGRP-null mice. (A) Schematic
`diagrams indicating the structures of ribonuclease protection probes containing adjacent intron /exon sequences at the calcitonin (exon 4) and
`CGRP- specific (exon 5) splice junctions are shown for both wild -type and modified aCGRP (CGRP -null) alleles; the migration positions of
`predicted RNase protection fragments are indicated. (B) RNase protection analysis of total RNA isolated from wild -type ( + / +), heterozygous
`( + / -), or aCGRP-null ( -/ -) whole mouse brain is presented and the migration positions of expected RNase protection fragments are indicated.
`(C) RNase protection analysis of total RNA isolated from the thyroid gland is presented. A separate probe for [3-actin was included in analyses of
`thyroid RNA to determine the relative levels of calcitonin mRNA expression in wild -type ( + / +) and «CGRP -null ( -/ -) animals.
`
`

`

`Cardiovascular and NMJ Analysis of cx CGRP -Null Mice
`
`103
`
`reactivity (CGRP -LI) was observed in many CNS and
`peripheral regions including the lateral superior olive
`nucleus (Fig. 3A), the mesencephalic trigeminal nucleus
`(Fig. 3B), and pancreatic ganglia (Fig. 3C) from wild -
`type mice, while CGRP -LI was conspicuously absent
`from the analogous regions in aCGRP -null animals. In
`contrast, identical CGRP staining patterns were ob-
`served in the myenteric ganglia from both wild -type
`and mutant mice (Fig. 3D), where 13,-CGRP is known to
`be the predominantly expressed CGRP isoform (Mul-
`berry et al., 1988; Sternini and Anderson, 1992). Immuno-
`histochemical analysis of aorta segments with an antise-
`rum directed against a general neuronal marker (PGP
`
`wild -type
`
`aCGRP -null
`
`A
`
`lateral superior olive nucleus
`
`mesen cephalic trigeminal nucleus
`
`r reirrp
`
`,
`
`46'p
`.
`Jib
`
`.
`
`,
`
`pancreatic ganglia
`
`.f
`/.,
`a
`
`i
`
`N,
`
`,
`
`ra1
`
`myenteric ganglia
`
`Immunohistochemical localization of CGRP in tissues from
`FIG. 3.
`wild -type and «CGRP -null mice. Fixed and paraffin- embedded tis-
`sues were prepared and 4 -pm sections were labeled using an amide -
`specific CGRP antiserum and counterstained with Harrison's hema-
`toxylin as described under Experimental Methods. Arrows indicate
`the presence of a positive immunoperoxidase reaction product. Magni-
`fication, 250X (A, B); 400x (C, D).
`
`+1+
`
`-I-
`
`FIG. 4. Analysis of CGRP -positive perivascular innervation in the
`aorta of wild -type ( + / +) and mutant ( -/ -) mice. Whole -mount
`sections of mouse aorta were labeled using an antiserum directed
`against either a general neuronal /neuroendocrine marker (PGP9.5)
`(a,b) or CGRP (a , b ) and a Cy- 3- labeled fluorescent second antibody.
`Magnification, 100x.
`
`9.5; Thompson et al., 1983) detected a rich network of
`perivascular innervation (Figs. 4a and 4b)
`in both
`wild -type and mutant animals. The presence of CGRP -
`positive nerve fibers, however, was distinctly absent
`from CGRP -null mice (Fig. 4b ), suggesting that a
`majority of the CGRP -positive perivascular fibers solely
`contain the «CGRP peptide. Since previous studies have
`demonstrated that CGRP is coreleased with acetylcho-
`line at the neuromuscular junction (NMJ) (Sakaguchi et
`al., 1991; Uchida et al., 1990, 1991), the presence of
`CGRP -LI was examined in skeletal muscles that were
`fixed, sectioned, and double -labeled with antibodies to
`CGRP plus rhodamine -a- bungarotoxin
`(rBTX) . BTX
`binds tightly and specifically to AChRs in the postsynap-
`tic membrane and thereby marks synaptic sites (Figs. 5a
`and 5b). In muscles from wild -type neonates, CGRP was
`readily detected at most NMJs (Fig. 5a ), while no
`immunoreactivity was observed at most synaptic sites
`in muscles from mutant neonates (Fig. 5b ). At a small
`minority (-10%) of mutant end -plates, however, very
`faint CGRP -LI was detectable (data not shown), suggest-
`ing that low levels of 13,-CGRP were present. No CGRP
`immunoreactivity was detected in wild -type or mutant
`NMJs from adult animals (data not shown), consistent
`with previous reports that CGRP levels are highest in
`embryonic or neonatal muscles and then often fall to
`undetectable levels in adulthood (Andreose et al., 1994;
`Matteoli et al., 1990a). These results demonstrate that
`CGRP expression was successfully deleted in regions
`where «CGRP is the predominantly expressed peptide
`isoform. The absence of «CGRP -LI in perivascular fibers
`and at neuromuscular end -plates further indicates that
`
`

`

`104
`
`expression of I3-CGRP was not upregulated to compen-
`sate for the loss of the a- peptide in these areas.
`
`A
`
`200-
`
`Mil
`
`(+/+)
`(/)
`
`Luetal.
`
`- 750
`
`- 500
`
`-250
`
`m
`
`.E " 6 a) 1,75'
`,a, .
`
`Regulation of Basal and Exercise -Induced
`Cardiovascular Hemodynamics
`Intravenous administration of CGRP elicits marked
`hypotensive as well as positive inotropic and chrono-
`tropic cardiac responses in both rat and human subjects
`(Ando et al., 1990; Bell and McDermott, 1994; Gardiner et
`al., 1991). On the basis of potent vasodilator effects and
`the perivascular localization of CGRP (Brain et al., 1985;
`Poyner, 1992; Struthers et al., 1986; Tan et al., 1995), it has
`been postulated that CGRP plays a role in the regulation
`of blood pressure and regional organ blood flow both
`under normal physiological conditions and in the patho-
`physiology of hypertension (Edvinsson et al., 1989;
`Kawasaki et al., 1990) . To investigate whether the release
`of «CGRP from periarteriolar nerve fibers or CNS nuclei
`contributes to the regulation of systemic vascular resis-
`tance, basal blood pressure and heart rate were mea-
`sured in wild -type and aCGRP-null animals. Direct
`measurements were obtained from left carotid arterial
`lines in conscious, unrestrained, chronically instru-
`mented adult male mice from 8 to 16 weeks of age.
`
`150-
`
`100-
`
`50-
`
`0
`
`200-
`
`150-
`
`100-
`
`50-
`
`0
`
`B
`
`a)
`
`( + / +)
`
`( -I -)
`
`( +1 +)
`
`( -/ -)
`
`o
`
`MAP
`
`HR
`
`(+/+)
`(-I-)
`
`NM
`
`1 S
`
`( -I -)
`( +1 +)
`MAP
`
`( +1 +)
`
`HR
`
`*
`
`- 750
`
`-500
`
`- 2.50
`
`o
`
`a
`
`au
`
`C E ,
`
`@ Ç
`_o
`
`co
`
`(7)
`
`+/+
`
`-I
`
`o
`o
`
`FIG. 6. Comparisons between basal and exercise -induced cardiovas-
`cular function in wild -type and «CGRP -null mice. (A) Mean arterial
`pressure (MAP) and heart rate (HR) were determined using a left
`carotid cannula connected to a blood pressure transducer as described
`under Experimental Methods. Stable HR and BP, after 2 h of animal
`equilibration, were determined for wild -type (n = 12) and «CGRP -
`null (n = 14) mice; data represent mean values (±SEM) of 120 data
`samples taken over a 2 -min period. Heart rate, systolic blood pressure,
`and diastolic blood pressure values were 647.0 ± 16 beats /min (bpm),
`129.3 -!- 3.1 mm Hg, and 133.7 -!- 3.4 mm Hg for wild -type animals,
`respectively, versus 604.9 ± 25 bpm, 133.7 ± 3.4 mm Hg, and 114.7 ±
`2.9 mm Hg for aCGRP-null mice. (B) Mean arterial pressure and heart
`rate were determined during 3 min of swimming exercise for wild -
`type (n = 6) and «CGRP -null (n = 6) mice; data represent mean values
`( ±SEM) of 360 data samples taken over the entire 3 -min period.
`Systolic and diastolic blood pressure values were 157.7 ± 6.9 and 136.0 -!-
`5.9 mm Hg versus 155.6 ± 4.4 and 134.9 ± 5.8 mm Hg for wild -type
`and aCGRP-null animals, respectively. Heart rates were 679.4 ± 20.7
`and 604.9 ± 25 bpm for wild -type and aCGRP-null animals; *P < 0.05.
`
`FIG. 5. CGRP immunoreactivity is absent from the motor end -plates
`of neonatal aCGRP-null animals. Muscles from wild -type ( + / +) and
`«CGRP -null ( -/ -) neonates were fixed, permeabilized, and double -
`stained with rhodamine a- bungarotoxin (rBTX; a, b), which binds to
`AChRs and thereby marks synaptic sites, and antiserum to CGRP
`(a ,b ). CGRP immunoreactivity is abundant at normal end -plates, but
`undetectable at most mutant end -plates. Muscle fiber cytoplasm is not
`stained by either label and is therefore not visible in these micro-
`graphs. Scale bar is 10 pm.
`
`Control and «CGRP -null mice exhibited comparable
`cardiovascular values including mean arterial blood
`pressure and heart rate (Fig. 6A) under basal conditions
`after a 2 -h equilibration period in the laboratory environ-
`ment. While basal systolic pressure was indistinguish-
`able between wild -type and mutant animals, there was
`a slight decrease in basal diastolic pressure in CGRP -
`null mice.
`
`

`

`Cardiovascular and NMJ Analysis of et CGRP -Null Mice
`
`105
`
`While the peripheral vasodilatory and central pressor
`activities of CGRP did not appear to contribute to the
`regulation of systemic vascular resistance and sympa-
`thetic outflow under basal conditions, these activities
`may be critical during a physiological stress such as
`muscular exertion. Previous studies have been contradic-
`tory regarding alterations in circulating CGRP levels in
`response to brief maximal exercise (Brooks et aI., 1990b;
`Schifter et al., 1995), although CGRP has been demon-
`strated to affect skeletal muscle regarding nicotinic
`receptor function (Mulle et al., 1988; New and Mudge,
`1986) and the activity of the Na + /K+ pump (Andersen
`and Clausen, 1993; Clausen and Nielsen, 1994) . To
`further examine potential differences in cardiovascular
`function between wild -type and «CGRP -null animals,
`mice were subjected to a 3 -min period of swimming
`exercise. Although the expected increases in exercise -
`induced cardiovascular values were readily observed
`(Fig. 6B), there was only a small difference in the mean
`heart rate measured between the two groups and no
`significant differences in other cardiovascular param-
`eters were noted. These results indicate that the ablation
`of «CGRP expression in mice does not lead to gross
`alterations in the regulation of cardiovascular homeosta-
`sis, indicating that «CGRP is not required for the
`systemic regulation of blood pressure and heart rate.
`
`Analysis of Autonomic Cardiovascular Regulation
`The autonomic nervous system can affect systemic
`cardiovascular regulation by redistributing blood flow
`to different areas of the body, by increasing the pumping
`activity of the heart, and, especially, by providing rapid
`control of arterial pressure via peripheral noradrenergic
`vasoconstriction (Dampney, 1994) . The lack of signifi-
`cant measurable difference in basal or exercise -induced
`cardiovascular function in «CGRP- deficient animals
`could reflect compensatory responses by alterations in
`sympathetic tone and /or autonomic responsiveness.
`The extent of autonomic contribution to the regulation
`of cardiovascular hemodynamics was therefore deter-
`mined in both wild -type and aCGRP -null mice subse-
`quent to ganglionic blockade. In both groups of mice,
`intravenous administration of the ganglionic blocker
`pentolinium resulted in a rapid decrease in MAP and
`HR that reached maximal, stable values approximately
`10 min after drug infusion (Fig. 7A) . Maximal decreases
`in MAP and HR were 37.7 ± 1.4% versus 38.2 ± 3.3%
`and 45.5 ± 2% versus 40.0 -!- 2.9% of basal values for
`control and mutant animals, respectively (Figs. 7A and
`7B), indicating no significant difference in the auto-
`
`nomic contribution to systemic cardiovascular hemody-
`namics.
`Although ganglionic blockade can allow an assess-
`ment of the absolute level of autonomic regulation, it
`does not address the rate of responsiveness of the
`autonomic nervous system to perturbations in cardiovas-
`cular parameters. To examine autonomic responsive-
`ness in control and mutant mice, a peripheral vasocon-
`strictor, phenylephrine (an al agonist), was infused
`intravenously and the rate of increase for the systolic
`blood pressure relative to the corresponding decrease in
`heart rate was quantified; this value is referred to as the
`baroslope and represents an index of baroreceptor reflex
`sensitivity (Cerutti et al., 1995). Analysis of wild -type
`and mutant animals revealed comparable baroslope
`values (Figs. 7C and 7D), indicating that the chronic and
`systemic absence of a potent vasodilator such as «CGRP
`in mutant animals did not lead to an adaptation or
`"resetting" of the baroreceptor response.
`
`Renin - Angiotensin Regulation of Systemic
`Cardiovascular Function
`The renin -angiotensin system (RAS) has important
`influences upon cardiovascular hemodynamics based
`largely upon the properties of angiotensin II (Ang II) as
`a potent vasoconstrictor and an aldosterone- stimulating
`peptide (Davis and Roberts, 1997). Angiotensin convert-
`ing enzyme not only generates the vasopressor peptide
`Ang II, but also inactivates the vasodilator actions of
`bradykinin (Brooks et al., 1990a). The final component of
`the renin -angiotensin regulatory cascade is activation of
`the AT1- subtype of angiotensin receptor, leading to
`vasoconstriction, liberation of catecholamines from sym-
`pathetic nerve endings, and a corresponding increase in
`blood pressure (Davis and Roberts, 1997). The absence
`of alterations in systemic cardiovascular regulation in
`animals deficient in «CGRP could reflect compensatory
`decreases in the level of RAS activity. To test this
`hypothesis, the basal contribution of RAS in the regula-
`tion of cardiovascular hemodynamics was assessed by
`intravenous administration of losartan (Merck -Du-
`Pont), an AT1 receptor antagonist. AT1 receptor blockade
`elicited similar cardiovascular responses from wild -type
`and aCGRP -null animals ( +/+ mice, MAP = 103.5 ± 4.5
`mm Hg, or 18.0 ± 2.6% decrease, n = 6; -/ - mice,
`MAP = 94.3 ± 6.2 mm Hg, or 22.2 -!- 4.4% decrease,
`n = 7). Since the extent of blood pressure reduction after
`AT1 receptor blockade was roughly equivalent in both
`groups of animals, these results indicate that significant
`alterations in RAS activity had not occurred in homozy-
`
`

`

`106
`
`A
`
`200-
`
`o
`0
`150-
`Hmip
`
`100-
`
`o
`
`o ¡° ::°OpOiiOD
`0 .,.. Yt! im
`ooó000" Qi_ °- "
`
`+1+
`
`-/-
`
`- 750
`
`C
`
`Lu et al.
`
`HR
`MAP
`
`o
`
`- 500
`
`- 250
`
`o
`
`10
`
`20
`Time (min)
`
`30
`
`40
`
`B
`
`150-
`
`_ (+1+)
`
`(-/-)
`
`I
`
`1
`
`o 1 ( + / +)
`
`( -/ -) 1 ( + / +)
`
`750
`
`500
`
`D
`
`a
`Q
`o
`
`o
`
`(D
`_a
`
`1.5-
`
`0.5-
`
`\ systolic blood pressure
`(mm Hg)
`
`mom
`
`1
`
`(+/+)
`(-/-)
`
`(+1+)
`
`( -I -)
`
`MAP
`HR
`FIG. 7. Autonomic regulation of cardiovascular function. Alterations in cardiovascular response were assessed by monitoring HR and MAP of
`wild -type and aCGRP -null mice after intravenous administration of the ganglionic blocker pentolinium (10 mg /kg body weight). (A) A
`representative time course is presented comparing alterations in MAP and HR over a 35 -min period. Data were sampled every 0.5 s and data
`points represent the mean value for each minute. (B) A summary of stabilized MAP and HR in wild -type and aCGRP -null mice after pentolinium
`infusion is presented; in general, stabilized cardiovascular parameters were achieved 15 -20 min after drug administration. (C) Analysis of
`baroreceptor reflex sensitivity in wild -type and aCGRP -null mice was assessed after intravenous administration of phenylephrine, a peripheral
`vasoconstrictor as described under Experimental Methods. During the linear phase of the increase in systolic blood pressure, alterations in the
`heart beat duration (L R -R interval) were monitored. Best -fit lines for determination of baroreceptor slope were determined by linear regression
`analysis using GraphPad Prism (GraphPad Software). A typical baroslope response is presented for wild -type and aCGRP -null animals. (D) A
`summary of derived baroslopes for wild -type (n = 6) and aCGRP -null (n = 8) mice is presented; all data represent mean values ( ±SEM).
`
`gous mutant mice to compensate for the developmental
`absence of aCGRP.
`
`Neuromuscular Junction Morphology
`Numerous studies have suggested that CGRP may
`exert a modulatory or trophic effect at the neuromuscu-
`lar junction based upon its localization and corelease
`with acetylcholine (Sakaguchi et al., 1991; Sala et al.,
`1995) and its effects in regulat