Regulation of the LIM-type homeobox gene islet-1 during neuronal regeneration

Pergamon

PII:

Neuroscience Vol. 88, No. 3, pp. 917–925, 1999 Copyright  1998 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/99 $19.00+0.00 S0306-4522(98)00263-2

REGULATION OF THE LIM-TYPE HOMEOBOX GENE ISLET-1 DURING NEURONAL REGENERATION E. M. HOL,*§ F.-W. SCHWAIGER,*‡ A. WERNER,* A. SCHMITT,*† G. RAIVICH,* and G. W. KREUTZBERG* *Department of Neuromorphology, Max-Planck-Institute of Neurobiology, D-82152 Martinsried, Germany †Department of Neurology, Technical University, School of Medicine, Aachen, Germany Abstract––Peripheral nerve lesion leads to prominent changes in gene expression in the injured neurons, a process co-ordinated by transcription factors. During development the transcription factor islet-1 plays an important role in differentiation and axogenesis. In axotomized adult neurons a process of axonal regrowth and re-establishment of the neuronal function has to be activated. Thus, we studied changes in the expression of islet-1 after axotomy, under the assumption that frequently developmentally regulated factors are reactivated during neuronal regeneration. We investigated the regulation of islet-1 expression with (i) semi-quantitative reverse transcription polymerase chain reaction and (ii) confocal microscopy in combination with quantitative image analysis. Islet-1 expression was suprisingly down-regulated in motoneurons and sensory neurons of adult rats after axotomy. A maximal reduction in the expression level was reached between day 3 and 7 after nerve lesion, a period of extensive axonal sprouting. Islet-1 expression attained control level at day 42 after lesion, a time-point at which target reinnervation takes place. The decreased expression of islet-1 during axonal regeneration is in contrast to the high levels of islet-1 expression during axogenesis in the developing nervous system. Thus, the proposed role of islet-1 in axonal target finding during axogenesis could not be confirmed in the adult rat. The observed down-regulation of islet-1 rather suggests that the activation of downstream genes important for the embryonic pattern of axonal path finding is suppressed. Moreover, in the adult nervous system islet-1 might be one of the transcription factors regulating the expression of proteins significant for the physiological intact neuronal phenotype.  1998 IBRO. Published by Elsevier Science Ltd. Key words: motoneurons, sensory neurons, transcription factor, LIM-type homeobox genes, peripheral nerve lesion, axotomy.

Following axonal lesion motoneurons basically possess the capacity of regeneration. They grow new axons, which leads to reinnervation of the peripheral target tissue and functional restitution. During this process profound changes in the morphology, the metabolism and the regulation of gene expression in the motoneurons have been observed (for review see Refs 12 and 21). The molecular mechanisms underlying successful regeneration are probably regulated by changes in the expression and activation level of transcription factors. Previous studies have indeed shown that peripheral nerve injury results in the induction of immediate-early genes such as c-jun14,17 ‡To whom correspondence should be addressed. §Present address: Netherlands Institute for Brain Research, 1105 AZ Amsterdam, The Netherlands. Abbreviations: BSA, bovine serum albumin; cDNA, copy DNA; DRG, dorsal root ganglion; EDTA, ethylenediaminetetra-acetate; FITC, fluorescein isothiocyanate; GAP-43, growth-associated protein-43; IR, immunoreactivity; OLV, optical luminosity value; PBS, phosphate-buffered saline; PCR, polymerase chain reaction; RISC, relative intensity of staining coefficient; RT, reverse transcription. 917

and junD15 in sensory and motoneurons. These immediate-early genes are known to encode transcription factors (reviewed in Ref. 25). In addition, it has been reported that the POU-domain transcription factor Oct-2 is up-regulated in sensory neurons after axotomy.4 There are indications of developmental phenotype being reinduced in mature neurons undergoing regeneration.3 Thus it seems possible that transcription factors responsible for the control of neuronal development and the establishment of axonal connections are reactivated postlesionally as part of a genetic regeneration program. Transcription factors belonging to one of the homeobox gene families appear to be promising candidates, as homeobox genes are involved in cellular patterning and phenotype determination.20 The LIM-type homeobox genes (islet-1,2 and lim-1,2,3) are of particular interest, since they are expressed in motoneurons and sensory neurons shortly after the last mitosis.10,34 It has been demonstrated that the islet-1 protein is still expressed in a subset of motoneurons in the spinal cord and brainstem and in the sensory neurons of the

918

E. M. Hol et al.

adult rat.32 The expression of various combinations of LIM-type homeobox genes in a neuron, the LIMcode, determines most probably in part the final identity and function of the different motoneuron populations in the developing spinal cord.23,26,33 In addition, has it been suggested that the LIM-code defines subclasses of motoneurons to select distinctive axonal pathways to their peripheral targets during embryogenesis.34 Since regeneration implies regrowth of axons, target finding and restoration of the functional identity of the lesioned neurons, islet-1 could play a role in this process. Therefore, we investigated the regulation of islet-1 expression in motoneurons and sensory neurons after axotomy in adult rats. EXPERIMENTAL PROCEDURES

Animals and surgical procedures Adult male Wistar rats (BD-Mpi (25); two to three months old) were operated under deep ether anaesthesia. The right facial nerve was transected or crushed for 10 s with a fine patterned forceps at the level of the stylomastoid foramen. In a second group of rats the right sciatic nerve was either transected or crushed for 30 s with a haemostatic forceps at mid-femoral level.8 The rats were killed at zero, one, three, seven, 14, 23 and 42 days after operation, perfused for 10 min with 100 mM phosphate-buffered saline (PBS; pH 7.4) and investigated tissues [facial nuclei or dorsal root ganglia (DRG) L4, L5 and L6] were removed and immediately frozen. For RNA isolation the injured right facial nuclei of two animals each were punched out and pooled. In addition, the right DRGs (L4, L5 and L6) from a single animal were pooled. The left facial nuclei and the left DRGs (L4–L6) served as a negative control. RNA isolation and copy DNA synthesis The facial nuclei were homogenized with a Branson sonifier B12 at 100 W for 30 s and the DRGs with an Ultra-Turrax T8 (Janke & Kunkel, Germany) at maximum speed for 15 s. Total RNA was isolated using TRISOLV (Biotecx Laboratories, Houston, TX, U.S.A.), a guanidinium isothiocyanate/phenol-based extraction solution, according to the manufacturer’s instructions. The RNA pellet was dissolved in 20 µl diethylpyrocarbonate-treated distilled water. Single-stranded copy DNA (cDNA) was synthesized from either a quarter of the total RNA extracted from two pooled facial nuclei or 500 ng total RNA from DRGs according to the manufacturer’s manual (Unites States Biochemical, Cleveland, OH, U.S.A.). The cDNA was stored at
80C until further use. Semi-quantitative reverse transcription–polymerase chain reaction All polymerase chain reactions (PCRs) were carried out in a final volume of 12.5 µl, containing 1PCR buffer [16 mM (NH4)2SO4, 67 mM Tris–HCl (pH 8.8 at 25C), 0.1% Tween-20; Eurobio, Les Ulis, France], 1.5 mM MgCl2 (Eurobio), 200 µM dNTPs (MBI Fermentas, Vilnius, Lithuania), 1 µCi (=13 nM) [á-32P]dATP, 1 µCi (=13 nM) [á-32P]dCTP (Amersham, Buckinghamshire, U.K.), 1 µM sense and 1 µM antisense primer. The PCR mixes were covered with one drop of mineral oil. Before the hot start of the PCR 1 µl of the 1:10 diluted cDNA samples was added to each PCR tube. The PCR was started at 94C with 1.25 U Taq DNA-polymerase (Eurobio) together with the radionucleotides.

Islet-1 was amplified with the sense primer 5 AAACTAATATCCAGGGGATGACAGG-3 starting at position 766 and the antisense primer 5 -CTCAGTA CTTTCCAGGGCGG-3 starting at position 889 of the rat islet-1 mRNA sequence (EMBL accession code: S69329), resulting in a DNA product of 123 bp in length. The growth-associated protein-43 (GAP-43) was amplified with the sense primer 5 -AGAAAGCAGCCAAGCTGAGGA GG-3 starting at position 636 and the antisense primer 5 -GACAGGGTTCAGGTGGGGG-3 starting at position 795 of the rat GAP-43 mRNA sequence (EMBL accession code: M16228), resulting in a DNA product of 159 bp in length. Cylophilin A was amplified with the sense primer 5 -CAACTCTAATTTCTTTGACTTGCGGG-3 starting at position 530 and the antisense primer 5 -AGAGATT ACAGGGTATTGCGAG-3 starting at position 635 of the rat cyclophilin A mRNA sequence (EMBL accession code: M19533), resulting in a 105 bp DNA fragment. The PCRs were carried out in a Biometra Trio-Thermoblock (Biometra, Go¨ttingen, Germany). For each experiment and cDNA synthesis the number of PCR cycles for optimal PCR conditions was determined independently for each primer pair to obtain logarithmic increase of PCR product (Fig. 1). The cycle temperature profiles were as follows: ‘‘islet-1’’ 30 s at 94C (with a 1 s time increment for each subsequent cycle), 20 s at 60C, 40 s at 74C for 32–35 cycles, GAP-43 30 s at 94C (with a 1 s time increment for each subsequent cycle), 10 s at 59C, 40 s at 74C for 25–27 cycles and ‘‘cyclophilin A’’ 30 s at 94C (with a 1 s time increment for each subsequent cycle), 10 s at 59C, 40 s at 74C for 23–25 cycles. After the final cycle the extension time was prolonged for 5 min in each PCR and then the reactions were soaked at 4C until further analyses. Next, 3.5 µl of the reactions was loaded on a 9% non-denaturing polyacrylamide gel (Rotiphorese gel 40, Roth, Karlsruhe, Germany) prepared in TBE buffer [90 mM Tris-(hydroxymethyl)aminomethan, 89 mM boric acid, 2 mM EDTA, pH 8]. The gels were dried and exposed to imaging plates (BAS-IIIs, Fuji Photo Film C, Kanagawa, Japan) for 40 min. Imaging plates were scanned in a phosphoimager (Fujix, BAS1000) and the signal was analysed and quantified with the TINA program (version 2.09 Raytest, Germany). The housekeeping protein cyclophilin A was used as an internal reference and the signal for islet-1 and GAP-43 was normalized for the cyclophilin A signal, as described earlier.11,29 PCR products of all investigated genes were cloned and their identity was confirmed with sequencing. For each time-point cDNA was analysed from three to six rats in triplicate. No signal was obtained when the PCR was performed on the RNA samples before cDNA synthesis, indicating that the isolated RNA was not contaminated with genomic DNA. The number of cycles resulting in optimal PCR conditions was chosen in the exponential phase of the PCR. The number of PCR cycles was 23 cycles for cyclophilin A, 32 cycles for islet-1 and 25 cycles for GAP-43. Immunohistochemistry Cryostat sections (20 µm) were postfixed in 3.7% formaldehyde (Merck, Darmstadt, Germany) in 10 mM PBS (pH 7.4) for 5 min, followed by 2 min each in 50% v/v acetone/H2O, 2 min 100% acetone and 2 min 50% v/v acetone/H2O at room temperature. The sections were washed twice in PBS and treated with 6.25 U/ml RNAse H (MBI Fermentas, Vilnius, Lithuania)/2 g/ml RNAse A (Boehringer Mannheim, Mannheim, Germany) in PBS at room temperature for 15 min. After preincubation with 5% donkey serum (Sigma, St Louis, MO, U.S.A.) in PBS at room temperature for 30 min, the sections were washed twice with PBS and were incubated overnight at 4C with a mouse monoclonal [clone 39.4D5, Developmental Studies Hybridoma Bank; 1:400 dilution in PBS/bovine serum albumin (BSA)] or a polyclonal antibody (K5, kindly provided

Islet-1 expression in regenerating neurons

919

Fig. 1. Semi-quantitative RT–PCR analysis of islet-1 expression. The curves show the logarithmic amplification phases of the cyclophilin A, islet-1 and GAP-43 PCR. The arrows indicate the optimal number of cycles used for further analysis of the cDNA samples. The same cDNA was also used for the experiment depicted in Table 1. The intensity of the PCR signal is analysed with a phosphoimager, the system output is defined as PSL-Bkg (photo-stimulated luminescence-background). (1) Control facial nucleus, (2) facial nucleus seven days after facial nerve transection, (3) control DRGs and (4) DRGs seven days after sciatic nerve transection. The arrows indicate the optimal number of PCR cycles used in subsequent experiments. by T. Jessel; 1:1000 dilution in PBS/BSA) against islet-1. Both antibodies have been reported to detect islet-1 as well as islet-2. Nevertheless, we were not able to detect any islet-2 expression in the here investigated tissues by applying reverse transcription (RT)–PCR using various different primer combinations (data not shown). The sections were washed twice for 5 min in PBS and for another 5 min in PBS/0.1% BSA (Sigma), followed by a 75 min incubation with a fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG secondary antibody (1:400 in PBS/BSA; Jackson Immunoresearch, West Grove, PA, U.S.A.) at room temperature. After three washes with PBS/BSA (each 5 min) the sections were incubated with a FITC-conjugated donkey anti-goat IgG (1:400 in PBS/BSA; Sigma) at room temperature for 75 min. The sections were washed twice with PBS/BSA, once with PBS and counterstained with a 1:10,000 dilution of TOPRO-3 iodide (Molecular Probes, Eugene, OR, U.S.A.) in PBS at room temperature for 45 min. After three final washes with PBS the sections were mounted with VECTASHIELD mounting medium (Vector Laboratories, Burlingame, CA, U.S.A.) and were scanned by confocal microscopy in combination with quantitative digital analysis. Confocal microscopy/quantitative fluorescence immunohistochemistry To quantify the islet-1 immunoreactivity in neuronal nuclei, digital micrographs of the FITC- and TOPRO-3fluorescence (10241024 pixels, 0–255/8-bit grey scale) were recorded simultaneously with a Leica TCS 4D confocal laser scanning microscope, 10 objective and constant Ar/Kr laser power using FITC (islet-1)- and Cy5 (TOPRO-3)-fluorescence channels. Fifteen consecutive equidistant levels were scanned per section (total vertical span 40 m), condensed to a 1-Mb TIFF-file for each fluorescence (I0.TIFF for islet- and T0.TIFF for TOPRO-3fluorescence) using the MaxIntens condensation algorithm. This algorithm selects the maximal intensity value for each pixel from the 15 available levels.

Quantification of islet-1 immunoreactivity was performed with a slight modification of the method described by Mo¨ller et al.24 Both condensed TIFF-files (T0.TIFF; I0.TIFF) were imported into the OPTIMAS 5.0 imaging analysis program and smoothed using a 100100 pixel box. In the first corrected TOPRO-3-file (T1.TIFF) the nonneuronal nuclei were eliminated by setting the threshold at mean +1.5-fold S.D. of the overall optical luminosity value of the TIFF-file (M+1.5 S.D. OLV), and resetting the OLV values of pixels above threshold to 0. To further differentiate between adjacent neuronal and non-neuronal fluorescence, a second TOPRO-3-file (T2.TIFF) was created using the SOBEL spatial gradient filter (33 pixel box) and contrast-enhanced using a 128–255 window. Subtraction of T2.TIFF from T1.TIFF (T3.TIFF), effectively eliminated the TOPRO-3-fluorescence hallow around non-neuronal nuclei which was still present in the T1.TIFF-files. Neuronal nuclei in this T3.TIFF-file were selected using M+1.5 S.D. threshold and these nuclear area profiles transferred to the smoothed islet-file (I0.TIFF). Mean OLV was determined for the whole I0.TIFF-file (OLVwhole) and over the neuronal nuclei (OLVnn). The relative intensity of staining coefficient for islet-1 immunoreactivity (islet-1-RISC) was calculated as logarithm of OLVnn to OLVwhole, i.e. RISC=log(OLVnn/ OLVwhole). RESULTS

Semi-quantitative analysis of the reverse transcription– polymerase chain reaction signal The RT–PCR method was established to assess changes in islet-1 mRNA expression in facial nuclei and DRGs (Fig. 1). Cyclophilin A was used as an internal reference11,29 controlling for differences in the amount of mRNA pipetted in the RT reaction and for differences in the efficiency of the RT reaction. Cyclophilin A is an abundantly expressed

E. M. Hol et al.

mRNA coding for a housekeeping protein. To validate the PCR method we additionally determined the level of expression of the GAP-43 mRNA. GAP-43 expression is strongly induced in facial motoneurons after facial nerve axotomy30 and in DRG neurons after sciatic nerve axotomy.6,35 PCR conditions were optimized for each primer pair and for each cDNA preparation (Fig. 1). Fig. 2. PCR amplification of islet-1, GAP-43 and cyclophilin A cDNA obtained from control and operated facial nuclei at different timepoints after facial nerve transection. Islet-1, GAP-43 and cyclophilin A PCR was performed independently with cDNA from one animal per time-point.

920

Regulation of islet-1 messenger RNA after transection of the facial nerve Changes in the expression levels of the LIM-type homeobox gene islet-1 and the GAP-43 gene were analysed zero, three, seven, 14, 23 and 42 days after a unilateral facial nerve transection with RT–PCR. GAP-43 mRNA was significantly up-regulated from day 3 after axotomy onwards (Figs 2, 3B). The expression levels declined after 22 days and no significant difference between the facial nucleus of the axotomized side and the control side could be measured 42 days after axotomy. The maximum increase in GAP-43 expression of 766133% compared to control (meanS.E.M., n=4, t-test: P<0.01) was reached at seven days after facial nerve transection. Statistical analysis revealed a significant decrease in the expression of islet-1 mRNA at three, seven, 14 and 22 days after facial nerve axotomy. The amount of mRNA was reduced to 3911% compared to control (meanS.E.M., n=4, t-test: P<0.05) at day 3 and 369% compared to control (meanS.E.M., n=4, t-test: P<0.05) at day 7 after transection. The time-course was comparable to that of GAP-43 mRNA expression (Figs 2, 3A) with the most pronounced decrease at day 3 and 7 post-facial nerve transection. Regulation of islet-1 protein after transection of the facial nerve Fluorescence immunohistochemistry at day 7 postaxotomy of the facial nerve was carried out in order to study regulation of islet-1 on the protein level. Islet-1 immunoreactivity (IR) was restricted to the nuclei of the facial motoneurons (Fig. 4A, C). The IR did not disappear completely after axotomy. All motoneurons were still islet-1 positive after nerve lesion, although the staining was less intense compared to the control situation (Fig. 4D). In addition, some intensely stained neurons were even present after axotomy. By counterstaining the tissue sections with the nuclear dye TOPRO-3 (Fig. 4B, D) we were able to quantify the intensity of the islet-1 IR within the neuronal cell nuclei. The differences in TOPRO-3 (Fig. 4B, D) staining intensity and in nuclear size between the motoneuron and glial cell nuclei, allow us to prepare a mask for quantification of the islet-1 IR in the nuclei of the motoneurons. Quantification of the immunofluorescence signal with confocal microscopy in combination with image analysis

Islet-1 expression in regenerating neurons

921

Fig. 3. Time-course of (A) islet-1 and (B) GAP-43 mRNA expression after facial nerve transection. The signal of islet-1 and GAP-43 was determined by scanning phosphoimages of the PCR products and normalized by the signal obtained for cyclophilin A amplification for the cDNA from the respective facial nucleus. For each time-point the percentage increase or decrease compared to control facial nucleus was calculated. *P<0.05, **P<0.01 and ***P<0.001 two-tailed, paired Student’s t-test, n=3–8.

revealed a statistically significant (with P<0.01 in a two-tailed paired t-test) down-regulation of the islet-1 protein in the nuclei of the facial motoneurons (Fig. 4C). Seven days after axotomy the RISC decreased significantly from 0.500.02 (mean S.E.M., n=4) on the control side to 0.350.01 on the operated side (P<0.01; two-tailed paired Student’s t-test).

crush and to 403% (P<0.01) after transection. The level of expression of islet-1 mRNA in sensory neurons was decreased to 755% (P<0.01) compared to the control after sciatic nerve crush and decreased to 687% (P<0.05) compared to the control after sciatic nerve transection. Thus, no principal difference in the regulation of islet-1 and GAP-43 was obtained between the motoneurons and sensory neurons and between the lesion types.

Regulation of islet-1 in sensory neurons and motoneurons after crush and transection

DISCUSSION

The facial nucleus is a pure cholinergic motor nucleus, thus by transecting the facial nerve motoneuron regeneration is studied. To investigate whether the changes in islet-1 expression after nerve lesion are motoneuron specific, we compared the results obtained from facial nucleus with the islet-1 expression in the primary sensory neurons of DRG after lesion of the sciatic nerve (Table 1). Alterations in the expression level of islet-1 mRNA were evaluated with semi-quantitative RT–PCR on day 7 following the nerve lesion. Again, up-regulation of GAP-43 mRNA expression was used as a positive control. As compared to the facial nucleus a down-regulation of islet-1 was received. Furthermore, we compared islet-1 regulation in the transection model with the regulation in a crush model. After transection the nerve is separated in two parts in contrast to a nerve crush. Hence after nerve crush there is a continuous support by the Schwann cells which speeds up the re-growth of axons to reinnervate their targets. This might be reflected in different expression patterns of islet-1 in the two types of lesions. For this separate experiment in facial motoneurons, seven days after operation, the amount of islet-1 mRNA on the operated side was reduced to 545% (P<0.05) compared to control side after

We investigated the expression of the LIM-type homeobox gene islet-1 in two systems of peripheral nerve lesion to gain insight in the molecular mechanisms underlying the process of neuronal regeneration. Islet-1 was found significantly down-regulated in axotomized motoneurons on the transcriptional as well as on the protein level. Furthermore, we demonstrated that islet-1 mRNA was also down-regulated in axotomized sensory neurons. This report is the first to our knowledge that shows a dynamic change in the expression of a LIM-type homeodomain protein in regenerating neurons of adult rats. The homeodomain protein islet-1 plays an important role in the development of motoneurons and neurons of the DRGs.34 The exact molecular mechanisms by which islet-1 and other LIM-type homeodomain proteins act is not clear yet. The LIM-type homeobox proteins, including islet-1, contain a homeodomain and two cys-his rich LIM-domains, that show high similarity to GATA zinc-fingers.28 This implies that the LIM-domain may be involved in binding to specific nucleic acid sequences. On the other hand, it has been proposed that the LIM-domain can interact with other proteins.13 In the two neuronal lesions investigated here islet-1 IR was exclusively restricted to the neuronal nuclei,

922

E. M. Hol et al.

Fig. 4. Islet-1 IR (A and C) and counterstaining with TOPRO-3 (B and D) in the rat facial nucleus. TOPRO-3 stained the nuclei of motoneurons (the less intense large round nuclei) as well as the nuclei of the astrocytes and microglial cells (the smaller bright nuclei). Note that the islet-1 IR (A and C) is restricted only to the motoneuron nuclei. In the operated facial nucleus (C=seven days post OPeration) a clear reduction in islet-1 IR in nuclei of the axotomized motoneurons was seen compared to the non-operated facial nuclei (A=CONtrol). Furthermore, the TOPRO-3 staining in this figure clearly shows the massive increase in microglial cell number in the facial nucleus, induced by the axotomy of the facial (B=CON, D=7D OP). These glial cells were void of islet-1 IR. Two putative populations of islet-1positive neurons are indicated: weakly stained islet-1-positive motoneurons (arrow) and intensely stained islet-1-positive motoneurons (arrowhead). Scale bar=50 µm.

suggesting that islet-1 is a physiological active transcription factor in neurons of the adult rat. Islet-1 is expressed shortly after the final mitosis in moto- and sensory neurons.10 Expression of islet-1 in these neurons is essential for their differentiation, as

was shown by elimination of functional islet-1 protein by gene targeting, which resulted in a failure in the differentiation of spinal cord motoneurons and a deficit in the development of the DRGs.5 In the developing spinal cord it has been shown that a

Islet-1 expression in regenerating neurons

923

Table 1. The effect of a crush lesion or a transection lesion on the expression of islet-1 and growth-associated protein-43 in motoneurons of the facial nucleus and sensory neurons of the dorsal root ganglia seven days postlesion Crush

Sensory neurons Motoneurons

Axotomy

Islet-1

GAP-43

Islet-1

GAP-43

755%** 545%*

24731%*** 46944%***

687%* 403%**

25315%*** 58449%***

The values represent per cent operated to control side. *P<0.05, **P<0.01 and ***P<0.001 two-tailed, paired Student’s t-test.

combinatorial expression of islet-1, islet-2, lim-1 and lim-3 defines subclasses of motoneurons.34 In addition, it has been proposed that this LIM-code co-ordinates the distinct functional phenotypes of the subclasses of motoneurons in the spinal cord23,34 and participates in determining the early axonal projections during development and in the regulation of the neuron-specific neurotransmitters.7,33,34 Thus, islet-1 is probably involved in the maintenance of the specific phenotypes of moto- and sensory neurons, as its expression continues in these neurons at a lower level during adulthood.32 Moreover, from studies in Drosophila, evidence has been provided, that a single islet gene controls axon-pathfinding as well as neuronal phenotype determination, in this case the differentiation and outgrowth of dopaminergic and serotonergic interneurons.33 Axonal outgrowth during neuronal regeneration seems to mimic axogenesis during neuronal development, since many proteins expressed in developing neurons are reinduced during neuronal regeneration.3 Since islet-1 is strongly expressed in developing neurons,34 we expected an induction of islet-1 expression when neurons regrow their axon. However, the expression of islet-1 mRNA and protein is down-regulated after axotomy in neurons, implicating that in contrast to the role of islet-1 in axogenesis, this protein seems not to be involved in axonal regrowth after lesion in an adult rat. Moreover, as in the adult animal, expression of islet-1 seems to be important for the maintenance of the phenotypic identity of neurons down-regulation of islet-1 might be related to the phenotypic changes regenerating neurons. Following axotomy, two types of neuronal reactions occur: (i) the expression of proteins important for neuronal function and which are related to neurotransmission is down-regulated and (ii) the expression of structural proteins is up-regulated (reviewed in Ref. 21). As a consequence of injury, neurons lose their physiological function. Instead of transmitting signals to their target cells, they are fully committed to axonal regrowth and restoration of connection. The temporal pattern of the decrease in islet-1 expression coincides with the time when the injured neurons suppress their genetic equipment necessary for function. Genes after injury similarly regulated as islet-1 might be, directly or

indirectly, positively regulated downstream effector genes of islet-1 in the adult animal. Enzymes needed for neurotransmitter synthesis and degradation in motoneurons, acetyl cholinesterase or choline acetyl transferase31 are possibly such effector molecules. Comparing the changes in islet-1 expression in sensory or motoneurons during regeneration is suprising. Islet-1 mRNA is down-regulated in motoneurons as well as in sensory neurons after nerve lesion, suggesting that islet-1 is involved in the regulation of genes that are expressed by both types of neurons. However, the two neuron classes possess different sets of neurotransmission related molecules, implying that islet-1 might not be involved in the regulation of these genes in the adult animal. On the other hand, there is a remarkable overlap between the expression of islet-1 and tyrosine hydroxylase or serotonin in many endocrine and neuronal cells of vertebrates33. Production of these two proteins in interneurons of islet-1-deficient Drosophila seems to be directly depending on expression of islet in a certain subset of neurons.33 This implies that islet-1 might be part of the regulatory cascade for genes coding for neurotransmitter-related molecules. Moreover, it is very well possible that in addition to islet-1 other regulatory factors, which are important in the pathways synthesizing the appropriate neurotransmitters, are different in sensory and motoneurons. Such factors might be e.g., other LIM-type homeobox genes or the newly described factor NLI/LDB11,18 that is known to interact with the LIM-1 domain. Next to an activating effect homeodomain proteins can also suppress the expression and translation of their downstream target genes.19,27,36 The temporal expression pattern of islet-1 mRNA coincides inversely, not only with that of genes related to axonal growth leading to repair, such as GAP-43, tubulin or actin,22,30 but also with the induction of some neuropeptides such as calcitonin gene-related peptide,2,9 galanin and vasoactive intestinal polypeptide.16 Nevertheless, the observed subpopulations of faint and strongly stained islet-1-positive neurons on the operated side do not overlap with the subpopulations defined by galanin or calcitonin gene-related peptide immunoreactive staining (data not shown). Many regeneration associated genes in the injured

924

E. M. Hol et al.

neurons are maximally induced between three and seven days after axotomy, coinciding with the minimum in islet-1 expression. The restoration of the basal level of islet-1 expression commences at the time-point when the axonal sprouts reinnervate their target tissue and is completed as functional repair and restoration of neuronal phenotype is reached. As deficiency of islet-1 protein indeed suppresses the transcription of neuropeptides,33 the downregulation of islet-1 might also be a crucial step in the induction of growth-associated proteins and neuropeptides. Further research must reveal whether there is a direct or indirect regulation of the transcriptional activity of regeneration-associated genes by islet-1. CONCLUSIONS

Down-regulation of islet-1 is a general phenomenon induced by axotomy (crush or transection lesion) in sensory and motoneurons. In the isletdeficient Drosophila axonal growth of motoneurons is not completely inhibited but appears to be highly disorganized and show defects in target selection.33 Although there is only indirect evidence of a similar regulation of genes involved in target finding by islet-1 in vertebrates, it might be possible that activation of such embryonally important factors as islet-1 during neuronal regeneration in the adult rat

interfere with the axonal guidance by the e.g., Schwann cells and target-derived neurotrophic factors. Hence down-regulation of islet-1 after nerve injury might be a basic event for the neuron to use the right set of genes for successful regeneration. This would suggest that the pathways promoting axonal outgrowth, during development and during regeneration are distinct. In conclusion, our data on the down-regulation of islet-1 in neurons during regeneration and the re-establishment of the basal islet-1 expression at the time-point of reinnervation of the target indicates that islet-1 might be of importance for the maintenance and recovery of the neuronal phenotype. A role for islet-1 in the process of axon pathfinding might be also postulated as it could be involved in the regulation of the embryonal genes responsible for axonal guidance.

Acknowledgements—We thank Anja Wo¨ppel and Maria Koch for excellent technical assistance, Karin Bru¨ckner and Dr James Chalcroft for expert help in photographical and digital documentation and Dr Gerhard Hager and Leonard Jones for discussion. The anti-islet-1 monoclonal antibody 39.4D5 developed by TM. Jessell was obtained from the Developmental Studies Hybridoma Bank maintained by the Department of Pharmacology and Molecular Sciences, John Hopkins University School of Medicine, Baltimore, MD 21205, U.S.A. and the Department of Biological Sciences, University of Iowa, Iowa City, IA 52242, U.S.A., under contract NO1-HD-2-3144 from the NICHD.

REFERENCES

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Agulnick A. D., Taira M., Breen J. J., Tanaka T., Dawid I. B. and Westphal H. (1996) Interactions of the LIM-domain-binding factor Ldb1 with LIM homeodomain proteins. Nature 384, 270–272. Arvidsson U., Johnson H., Piehl F., Cullheim S., Ho¨kfelt T., Risling M., Terenius L. and Ulfhake B. (1990) Peripheral nerve section induces increased levels of calcitonin gene-related peptide (CGRP)-like immunoreactivity in axotomized motoneurons. Expl Brain Res. 79, 212–216. Aubert I., Ridet J.-L. and Gage F. H. (1995) Regeneration in the adult mammalian CNS: guided by development. Curr. Opin. Neurobiol. 5, 625–635. Begbie J. L., Wood J. N., Anderson P. N. and Latchman D. S. (1996) Specific up-regulation of the POU domain transcription factor Oct-2 following axotomy. Neurosci. Lett. 207, 183–186. Chen E. W. and Chiu A. Y. (1992) Early stages in the development of spinal motor neurons. J. comp. Neurol. 320, 291–303. Chong M. S., Fitzgerald M., Winter J., Hu-Tsai M., Emson P. C., Wiese U. and Woolf C. J. (1992) GAP-43 mRNA in rat spinal cord and dorsal root ganglia neurons: developmental changes and re-expression following peripheral nerve injury. Eur. J. Neurosci. 4, 883–895. Daston M. M. and Koester S. E. (1996) Transcriptional regulation of axon pathfinding. Neuron 17, 5–8. De Koning P., Brakkee J. H. and Gispen W. H. (1986) Methods for producing a reproducible crush in the sciatic and tibial nerve of the rat and rapid and precise testing of return of sensory function; beneficial effects of melanocortins. J. neurol. Sci. 74, 237–246. Dumoulin F. L., Raivich G., Streit W. J. and Kreutzberg G. W. (1990) Differential regulation of calcitonin gene-related peptide (CGRP) in regenerating rat facial nucleus and dorsal root ganglion. Eur. J. Neurosci. 3, 338–342. Ericson J., Thor S., Edlund T., Jessell T. M. and Yamada T. (1992) Early stages of motor neuron differentiation revealed by expression of homeobox gene islet-1. Science 256, 1555–1560. Estus S., Zaks W. J., Freeman R. S., Gruda M., Bravo R. and Johnson E. M. Jr (1994) Altered gene expression in neurons during programmed cell death: identification of c-jun as necessary for neuronal apoptosis. J. Cell Biol. 127, 1717–1727. Fawcett J. W. and Keynes R. J. (1990) Peripheral nerve regeneration. A. Rev. Neurosci. 13, 43–60. Feuerstein R., Wang X., Song D., Cooke N. E. and Liebhaber S. A. (1994) The LIM/double zinc-finger motif functions as a protein dimerization domain. Proc. natn. Acad. Sci. U.S.A. 91, 10,655–10,659. Haas C. A., Donath C. and Kreutzberg G. W. (1993) Differential expression of immediate early genes after transection of the facial nerve. Neuroscience 53, 91–99. Herdegen T., Fiallos-Estrada C. E., Schmid W., Bravo R. and Zimmerman M. (1992) The transcription factors c-JUN, JUN-D and CREB, but not FOS and KROX-24, are differentially regulated in axotomized neurons following transection of rat sciatic nerve. Molec. Brain Res. 14, 155–156.

Islet-1 expression in regenerating neurons 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

925

Ho¨kfelt T., Zhang X. and Hallin-Wiesenfeld Z. (1994) Messenger plasticity in primary sensory neurons following axotomy and its functional implications. Trends Neurosci. 17, 22–30. Jenkins R. and Hunt S. P. (1991) Long-term increase in the levels of c-jun mRNA and jun protein-like immunoreactivity in motor and sensory neurons following axon damage. Neurosci. Lett. 129, 107–110. Jurata L. W., Kenny D. A. and Gill G. N. (1996) Nuclear LIM interactor, a rhombotin and LIM homeodomain interacting protein, is expressed early in neuronal development. Proc. natn. Acad. Sci. U.S.A. 93, 11,693–11,698. Kallunki P., Jenkinson S., Edelman G. M. and Jones F. S. (1995) Silencer elements modulate the expression of the gene for the neuron-glia cell adhesion molecule, Ng-CAM. J. biol. Chem. 270, 21,291–21,298. Keynes R. and Krumlauf R. (1994) Hox genes and regionalization of the nervous system. A. Rev. Neurosci. 17, 109–132. Kreutzberg G. W. (1995) Reactions of the neuronal cell body to axonal damage. In The Axon: Structure, Function and Pathophysiology (eds Waxman S. G., Kocsis J. D. and Stys P. K.), pp. 355–374. Oxford University Press, New York. Kreutzberg G. W. (1996) Principles of neuronal regeneration. Acta. neurochir. 66, Suppl., 103–106. Lumsden A. (1995) A ‘‘LIM code’’ for motor neurons? Curr. Biol. 5, 491–495. Mo¨ller J. C., Klein M. A., Haas S., Jones L. L., Kreutzberg G. W. and Raivich G. (1996) Regulation of thrombospondin in the regenerating mouse facial motor nucleus. Glia 17, 121–132. Pennypacker K. R. (1995) AP-1 transcription factor complexes in CNS disorders and development. J. Florida med. Ass. 82, 551–554. Pfaff S. L., Mendelsohn M., Stewart C. L., Edlund T. and Jessell T. M. (1996) Requirement for LIM homeobox gene Isl1 in motor neuron generation reveals a motor neuron-dependent step in interneuron differentiation. Cell 84, 309–320. Rivera-Pomar R., Niessing D., Schmidt-Ott U., Gehring W. J. and Jackle H. (1996) RNA binding and translational suppression by bicoid. Nature 379, 746–749. Sa´nchez-Garcı´a I. and Rabbitts T. H. (1994) The LIM domain: a new structural motif found in zinc-finger-like proteins. Trends Genet. 10, 315–320. Semple-Rowland S. L., Mahatme A., Popovich P. G., Green D. A., Hassler G. Jr, Stokes B. T. and Streit W. J. (1995) Analysis of TGF- 1 gene expression in contused rat spinal cord using quantitative RT–PCR. J. Neurotrauma 12, 1003–1114. Tetzlaff W., Alexander S. W., Miller F. D. and Bisby M. A. (1991) Response of facial and rubrospinal neurons to axotomy: changes in mRNA expression for cytoskeletal proteins and GAP-43. J. Neurosci. 11, 2528–2544. Tetzlaff W. and Kreutzberg G. W. (1984) Enzyme changes in the rat facial nucleus following a conditioning lesion. Expl Neurol. 85, 547–564. Thor S., Ericson J., Bra¨nnstro¨m T. and Edlund T. (1991) The homeodomain LIM protein Isl-1 is expressed in subsets of neurons and endocrine cells in the adult rat. Neuron 7, 881–889. Thor S. and Thomas J. B. (1997) The Drosophila islet gene governs axon pathfinding and neurotransmitter identity. Neuron 18, 397–409. Tsuchida T., Ensini M., Morton S. B., Baldassare M., Edlund T., Jessell T. M. and Pfaff S. L. (1994) Topographic organization of embryonic motor neurons defined by expression of LIM homeobox genes. Cell 79, 957–970. Van der Zee C. E. E. M., Nielander H. B., Vos J. P., Lopes da Silva S., Verhaagen J., Oestreicher A. B., Schrama L. H., Schotman P. and Gispen W. H. (1989) Expression of growth-associated protein B-50 (GAP43) in dorsal root ganglia and sciatic nerve during regenerative sprouting. J. Neurosci. 9, 3505–3512. Welter J. F., Gali H., Crish J. F. and Eckert R. L. (1996) Regulation of human involucrin promoter activity by POU domain proteins. J. biol. Chem. 271, 14,727–14,733. (Accepted 27 April 1998)