Doublecortin is a highly valuable endogenous marker of adult neurogenesis in canaries (2024)

The publisher's final edited version of this article is available free at Brain Behav Evol

1. The neuroanatomical distribution of doublecortin expression does not match the pattern of neurogenesis in the canary brain

Vellema and colleagues report that DCX-expressing cells, based both on immunohistochemistry for the protein and in situ hybridization for the corresponding mRNA, are found throughout the canary brain, with the most pronounced expression in the telencephalon, predominantly in the mesopallium and caudal nidopallium, which corresponds indeed to brain areas where an active incorporation of new neurons is observed in adult birds based on other methods. They indicate that scattered labeled neurons were also found “occasionally” in other regions outside the telencephalon (e.g. ventral tegmental area, substantia nigra, locus ceruleus).

These are not new observations, as this broader distribution was already known from our previous immunohistochemical study in canaries [Boseret et al. 2007]. A similar pattern has also been described in other avian species, and mammalian studies have likewise described DCX in brain regions that are not normally considered neurogenic, but might well be upon closer examination [Ernst et al. 2014; Kokoeva et al. 2007]. These non-telencephalic DCX-expressing cells are rare compared to telencephalic populations, and staining in these populations is usually of a different nature than in the telencephalon: it is weaker and not as crisp (fuzzy) [Boseret et al. 2007].

It has been accepted that DCX in mammals is a marker of young neurons, but that it also labels some cells that are reorganizing their dendritic arbor (another form of plasticity that requires microtubule reorganization and thus DCX expression). It is thus possible that these DCX cells indeed do not represent young newborn neurons, but this conclusion cannot be firmly established at this time. Our current understanding of adult neurogenesis in avian and mammalian brains is incomplete, and adult neurogenesis may occur in currently unidentified locations [Ernst et al. 2014; Kokoeva et al. 2007]. A broader than expected distribution of neurogenesis in the canary brain is suggested by the fact that Vellema et al. (2014) detected cells labeled by bromodeoxyuridine (BrdU) in sub-telencephalic brain regions that are not thought to recruit adult-born neurons (their Figure 7B).

2. Seasonal changes and hormonal effects on doublecortin expression do not match previously described changes in neurogenesis

Vellema and colleagues claim that the pattern of DCX distribution is similar in males and females and does not vary across seasons (based on the two examined time points), except in HVC and area X. The authors quantified the area covered by DCX-immunoreactive material in some brain areas, but it is unclear how extensive this quantification was. It seems that quantification concerned only area X and surrounding tissue. Furthermore, the authors only report relative expression, using plus and minus signs, and discuss the labeling in HVC sub-regions qualitatively. Based on these data, they claim that changes in DCX expression in HVC and area X “did not correlate with known patterns of neuron recruitment”. Two comments are in order here.

First, neurogenesis in the songbird brain is highly variable and controlled by a multitude of factors (strain, sex, testosterone, photoperiod, singing activity, social environment [Nottebohm 2008]). The impact of these factors on different aspects of neurogenesis (proliferation at the ventricle, migration, recruitment, differentiation and survival of neurons) is still largely unknown. It is impossible to predict the actual differences in neurogenesis between groups in the Vellema et al. study because neurogenesis was not investigated in these different groups of birds (different stages in the annual cycle, males vs. females, testosterone-treated or not) by an independent method, such as BrdU incorporation. Claiming that DCX does not correlate with neurogenesis is, therefore, not justified.

Second, the limited quantitative estimates for area X did not take into account the morphology of labeled cells: Vellema and colleagues only measured the surface covered by immunoreactive material. There are two morphological types of DCX-immunoreactive cells: fusiform, mostly bi-polar cells are probably very young neurons that are still engaged in the radial migration to their final destination, and round multipolar cells are presumably older neurons that have begun their differentiation. The temporal changes in numbers of these two cell types are substantially different [Balthazart et al. 2008; Yamamura et al. 2011]. Therefore, conclusions based on analyses that do not differentiate between these cell types seem unjustified.

3. Doublecortin is expressed in neurons of up to one year of age

In a potentially important experiment, Vellema et al. (2014) injected a small number of male canaries with BrdU and collected their brains 38 days (n=4), 60 days (n=4) and 365 days (n=2) later to analyze the expression of DCX in BrdU-labeled neurons. It is unfortunate that no information on the physiological state of these adult canaries was presented, since neurogenesis could, for example, be quite different in photosensitive or photorefractory birds [Balthazart et al. 2008]. More importantly, the injection schedule varied between birds (6 injections 4 hours apart in one day or 3 injections per day during two days) and readers are not told whether these two types of injections were equally distributed across the 10 birds and 3 survival times. Thus comparisons here seem to concern birds coming from three different experimental protocols that may not be directly comparable. These two patterns of injections in different groups of birds potentially living in different endocrine or social conditions likely resulted in a different degree of initial labeling that could interfere with the conclusions drawn from results observed 38, 60 and 365 days later.

The authors conclude from this experiment that DCX is still expressed in some BrdU-positive neurons at 38 and 60 days and that “even one year after BrdU injections we were still able to detect neurons expressing both BrdU and DCX, albeit sporadically”. No numbers are associated with this last statement (only one such neuron is shown in a photomicrograph), and we are not told where these few neurons expressing DCX after one year are located. A few adult neurons are known to express DCX because they are undergoing a plastic reorganization of their dendritic arbor [Kremer et al. 2013]. The “sporadic” expression observed by Vellema et al. may be related to this dendritic reorganization. If it concerns just a small fraction of neurons, this observation has little significance.

The authors suggest that their observation of DCX expression in BrdU-positive neurons 38 and 60 days after BrdU injections is a major issue especially because the percentage of double-labeled cells decreased only slightly between these two time points in HVC, NCM and area X (their Figure 7A). The total numbers on which these percentages are calculated are not indicated, making it difficult to appreciate the significance of these percentages (the decreases are said to be non-significant but no statistical analysis is presented). If we consider the reported numbers as representative of the entire nuclei, they remain consistent with our previous observations. We found that 10 days after BrdU injections approximately 70% of DCX-positive fusiform cells are BrdU-positive and that this percentage decreases to 30% by 30 days post-injection [Balthazart et al. 2008]. At 30 days, the percentage of round DCX-positive cells increases from 5 to 25%. In addition, the percentage of BrdU-positive cells that express DCX does not decrease between 10 and 30 days post-BrdU injection (73.38±7.57 and 75.37±13.22 % respectively; Balthazart and Ball, unpublished calculations based on results presented in [Balthazart et al. 2008]). Most if not all DCX-positive cells in HVC are thus young neurons labeled by BrdU, even if 100% co-labeling cannot be observed due to limitations of the BrdU method [Barker et al. 2013; Taupin 2007]. The limited decrease between 38 and 60 days in the percentage of BrdU-positive cells that are simultaneously DCX-positive observed by Vellema et al. indicates that the duration of DCX expression in neurons in different parts of the brain should be investigated further. This observation does not, however, invalidate the use of DCX as a marker, since the initial degree of co-labeling in their study is unknown and was likely to differ between groups.

4. Doublecortin expression does not correlate with BrdU labeling equally throughout the brain

In a second analysis of the brains of the same BrdU-injected subjects, Vellema and colleagues estimated in several brain areas the ratio of BrdU-positive to DCX-positive cells by dividing the absolute numbers of these two cell types (their Figure 7B). This ratio decreased from 38 to 60 days (consistent with DCX labeling young neurons!) and varied from one brain region to another (from 12% in HVC to less than 2% in the medial septum and thalamic region DMA), which according to Vellema et al. indicates that DCX cannot predict neurogenesis.

However, the physiological meaning of this ratio is questionable as long as it is based on single-labeled material and estimated from tissue collected 38 or 60 days after the BrdU injections. The BrdU injections label cells only for a few hours after injection [Barker et al. 2013] and only during DNA replication, whereas DCX labels young neurons continuously. It is therefore quite logical that only a small fraction of DCX-positive cells contain BrdU after 38 or 60 days; most of them were born after the injections (or shortly before). In our previous study, we showed that, of the young (fusiform) neurons arriving in HVC 10 days post-injection, 72% contain BrdU, but 20 days later this percentage is already down to 30% [Balthazart et al. 2008]. The low percentages observed here are thus very consistent with our understanding of what DCX and BrdU are measuring in cells.

The differences between brain regions that the authors find surprising are completely expected: the migration, recruitment and survival of new neurons in different brain regions are obviously different and take place over different periods of time. In HVC, for example, one half of the new neurons die within a month [Kirn et al. 1999], and their survival depends on endocrine condition [Kirn et al. 1994; Kirn and Schwabl 1997; Rasika et al. 1994] and possibly social environment, neither of which were described in detail in the paper. The numerator of the BrdU/DCX ratio (number of BrdU-positive cells) is thus expected to differ from one brain area to another. The observations presented in Figure 7B are therefore consistent with the idea that DCX is a reliable marker of neurogenesis.

5. Doublecortin is expressed in adult neurons expressing the immediate early gene egr-1

The authors contend that DCX is expressed in differentiated neurons of the nidopallium, because the DCX-positive cells express the neuronal marker NeuN. One such neuron is illustrated in their figure 6. How frequent is this type of co-labeling? The authors give the figure of 41.3 ± 4.9% for the nidopallium, but what is the basis of this calculation? Percent of all cells? Was cell type considered (fusiform or round or both together)? How many birds and how many sections were investigated? Is this observation specific to the nidopallium, or can a similar observation be made in other brain areas? Contrary to what is stated in the Vellema paper, NeuN is not exclusively an adult neuron marker, at least in mammals. In the mammalian brain new neurons begin expressing NeuN before they stop expressing DCX [Kempermann et al. 2004][Ernst et al. 2014]. If the same is true in canaries, it would not be surprising to find the degree of colocalization observed by Vellema and colleagues.

These authors report that DCX is expressed in physiologically active neurons, as indicated by the expression of the immediate early gene egr-1. Once again, this observation raises several questions: Was this co-localization specific to the auditory regions (NCM)? What percentage of cells exhibited this pattern of double-labeling? What was/were the stimulus(i) that induced this expression? Young developing DCX-positive neurons must synthesize a large number of proteins during their differentiation, and it is quite possible that transcription of the corresponding genes is controlled by the immediate early gene egr-1. Some of these proteins may be necessary for neuronal maturation, since in mice inactivation of the egr-1 gene results in alterations of the neurogenesis process [Veyrac et al. 2013]. In the absence of experimental data demonstrating that egr-1 expression in DCX-positive cells in NCM correlates with neuronal activation by species-specific auditory stimuli, it is difficult to assign a specific meaning to the observation that some DCX-positive cells express egr-1.

6. Doublecortin is expressed in differentiated projection neurons

Finally, Vellema et al. (2014) report that DCX is expressed in neurons that have established functional connectivity and have, therefore, presumably been mature for a significant period of time. In three birds the authors labeled some DCX-expressing neurons in HVC by injecting a retrograde tracer into the target nucleus robustus arcopallialis (RA), but not into area X. These data are consistent with DCX being a marker of new neurons, since RA-projecting neurons in HVC are replaced in adulthood, whereas area X-projecting neurons are not.

Furthermore, based on available H³-thymidine or BrdU experiments, it seems that the HVC to RA connections form around two weeks after neurons become post-mitotic and full connectivity is achieved within a month [Kirn et al. 1999]. It is therefore no surprise that some DCX positive neurons in HVC could be retrogradely labeled from RA since DCX is still expressed one to two months after neurons become post-mitotic.

• Balthazart J, Boseret G, Konkle AT, Hurley LL, Ball GF. Doublecortin as a marker of adult neuroplasticity in the canary song control nucleus hvc. The European journal of neuroscience. 2008;27:801–817. [PubMed] [Google Scholar]
• Barker JM, Charlier TD, Ball GF, Balthazart J. A new method for in vitro detection of bromodeoxyuridine in serum: A proof of concept in a songbird species, the canary. PloS one. 2013;8:e63692. [PMC free article] [PubMed] [Google Scholar]
• Boseret G, Ball GF, Balthazart J. The microtubule-associated protein doublecortin is broadly expressed in the telencephalon of adult canaries. J Chem Neuroanat. 2007;33:140–154. [PMC free article] [PubMed] [Google Scholar]
• Ernst A, Alkass K, Bernard S, Salehpour M, Perl S, Tisdale J, Possnert G, Druid H, Frisen J. Neurogenesis in the striatum of the adult human brain. Cell. 2014;156:1072–1083. [PubMed] [Google Scholar]
• Kempermann G, Jessberger S, Steiner B, Kronenberg G. Milestones of neuronal development in the adult hippocampus. Trends Neurosci. 2004;27:447–452. [PubMed] [Google Scholar]
• Kirn J, O'Loughlin B, Kasparian S, Nottebohm F. Cell death and neuronal recruitment in the high vocal center of adult male canaries are temporally related to changes in song. Proceedings of the National Academy of Sciences of the United States of America. 1994;91:7844–7848. [PMC free article] [PubMed] [Google Scholar]
• Kirn JR, Schwabl H. Photoperiod regulation of neuron death in the adult canary. J Neurobiol. 1997;33:223–231. [PubMed] [Google Scholar]
• Kirn JR, Fishman Y, Sasportas K, Alvarez-Buylla A, Nottebohm F. Fate of new neurons in adult canary high vocal center during the first 30 days after their formation. The Journal of comparative neurology. 1999;411:487–494. [PubMed] [Google Scholar]
• Kokoeva MV, Yin H, Flier JS. Evidence for constitutive neural cell proliferation in the adult murine hypothalamus. The Journal of comparative neurology. 2007;505:209–220. [PubMed] [Google Scholar]
• Kremer T, Jagasia R, Herrmann A, Matile H, Borroni E, Francis F, Kuhn HG, Czech C. Analysis of adult neurogenesis: Evidence for a prominent “non-neurogenic” dcx-protein pool in rodent brain. PloS one. 2013;8:e59269. [PMC free article] [PubMed] [Google Scholar]
• Nottebohm F. The discovery of replaceable neurons. In: Zeigler HP, Marler P, editors. Neuroscience of birdsong. Cambridge: Cambridge University Press; 2008. pp. 425–448. [Google Scholar]
• Rasika S, Nottebohm F, Alvarez-Buylla A. Testosterone increases the recruitment and/or survival of new high vocal center neurons in adult female canaries. Proceedings of the National Academy of Sciences of the United States of America. 1994;91:7854–7858. [PMC free article] [PubMed] [Google Scholar]
• Taupin P. Brdu immunohistochemistry for studying adult neurogenesis: Paradigms, pitfalls, limitations, and validation. Brain Res Rev. 2007;53:198–214. [PubMed] [Google Scholar]
• Vellema M, Hertel M, Urbanus SL, Van der Linden A, Gahr M. Evaluating the predictive value of doublecortin as a marker for adult neurogenesis in canaries (serinus canaria) The Journal of comparative neurology. 2014;522:1299–1315. [PubMed] [Google Scholar]
• Veyrac A, Gros A, Bruel-Jungerman E, Rochefort C, Kleine Borgmann FB, Jessberger S, Laroche S. Zif268/egr1 gene controls the selection, maturation and functional integration of adult hippocampal newborn neurons by learning. Proceedings of the National Academy of Sciences of the United States of America. 2013;110:7062–7067. [PMC free article] [PubMed] [Google Scholar]
• Yamamura T, Barker JM, Balthazart J, Ball GF. Androgens and estrogens synergistically regulate the expression of doublecortin and enhance neuronal recruitment in the song system of adult female canaries. J Neurosci. 2011;31:9649–9657. [PMC free article] [PubMed] [Google Scholar]