ADVANCING METHODS FOR STUDYING THE NEUROSCIENCE OF SEX VARIABILITY AND DIVERSITY

Sex variability and diversity are pervasive across taxa and genera. Society is heavily structured around sex/gender conceptualizations, and neuroscience research actively shapes these constructs. This dissertation adopts frameworks that challenge ideologies perpetuating sex/gender injustices and developed neuroscience-related methodologies that could advance the conceptualisations of sex. First, I studied B6D2F1 mice which exhibit sexual behaviour months after castration. This challenged the dogma that sexual behaviour is dependent on gonadal steroid hormones. I enhanced the behavioural analysis of these mice using deep learning techniques. Second, applying similar deep learning methods, I refined the behavioural analysis of Astatotilapia burtoni cichlids, which exhibit sexual behaviour in groups. Traditionally, neuroscience studies on sexual behaviour involve animal dyads singly isolated until experimentation. Third, recognising that laboratory cultivation alters cichlid behaviour and brain compared to the wild, I developed a framework utilising camera trap datasets to facilitate future neuroscience studies in wildlife settings. Lastly, in addition to sex at the behavioural level, numerous variables at cellular, molecular, and genetic levels are related to sex. Neuroscience currently lacks models to adequately construct human sex at these levels. To address this gap, I engineered a cell type relevant to human sex from stem cells. These efforts encompass various aspects of sex variability and diversity, emphasizing rigorous, precise, and transparent science to promote sex/gender justice and advance our construction of sex.

1. Advancing Methods for
Studying the Neuroscience of
Sex Variability and Diversity
Vic Shao-Chih Chiang 蔣紹志
July 2024
1

2. 2
B Velocci (2024) Cell
Individuals don’t fulfil the archetypes of all variables that relate to sex

3. 3
KO Smiley, SD Sun (2024) Horm Behav
Deconstructing sex: non-binary and non-essentialist

4. 4
MG Massa, K Aghi, MJ Hill (2023) Horm Behav
Science actively constructs the world that affected
lives of sex/gender diverse individuals

5. 5
Presentation Outline

6. 6
Dissertation Aim A: Background

7. 7
Sebocytes are important for sex & neuroscience
RW Clayton et al.R Paus (2020) Biol Rev
α-MSH, α-melanocyte-stimulating hormone
ACTH, adrenocorticotropic hormone
CRH, corticotropin-releasing hormone
CRHR-1, CRH receptor 1
CRHR-2, CRH receptor 2
GHR, growth hormone receptor
IGF-1, insulin-like growth factor-1;
IGF-1R, IGF-1 receptor
MC1R, melanocortin 1 receptor
MC2R, melanocortin 2 receptor
MC5R, melanocortin 5 receptor
POMC, proopiomelanocortin
PRL, prolactin; PRLR, prolactin receptor

8. 8
•non-invasive biological samples
•differentiate into any cell types
•retain the genetics of the donor
A Ray et al. RP Thummer (2021) Stem Cell Rev Rep
PBMC, peripheral blood mononuclear cells
Induced-pluripotent stem cells (iPSCs)

9. 9
Aim: engineer sebocytes from
induced pluripotent stem cells

10. 10
Directly activating transcription factors to drive cell differentiation
hiPSC, human induced pluripotent stem cells
TF, transcription factors
FACS, fluorescence-activated cell sorting
TRA1-60, T cell receptor alpha locus
scRNA-seq: single-cell RNA sequencing

11. 11
Activating combinations of transcription factors drove
cell differentiation (cells low in pluripotency)
0-10
3
10
3
10
4
10
5
PE-A
0
50
100
150
Count
Subset Name Count
TRA1-60low 4075
Singlets2 13581
PE, Phycoerythrin
TRA1-60, T cell receptor
alpha locus

12. 12
Few differentiated cells mapped to sebocytes transcriptomically
Endo, endothelial cells
Fibro, fibroblasts
KC, keratinocytes
T, T cells
VSMC, vascular smooth muscle
cells
B, B cells
UMAP, Uniform Manifold
Approximation and Projection

13. 13
Conclusions & Implications
•Outcome 1: generated differentiated cells
•Outcome 2: few differentiated cells resembled sebocytes
•Significance: study cell-level human sex & interactions with neuroscience
•Future Application: improve gender-affirming care

14. 14
Dissertation Aim B: Background

15. 15
Rodent sexual behaviour
Adapted from LP Yin & DY Lin (2023) Horm Behav

16. 16
Sexual behaviour may not depend on gonadal steroid
hormones under certain circumstances in many species
A Manning & ML Thompson (1976) Anim Behav
B6D2F1 mice with
testicles removed
B6D2F1 mice with
testicles intact

17. 17
Limitations with manual behavioural analysis
K Luxem et al. M Laubach (2023) Elife
•Time- consuming
•Labour- intensive
•Variation in analysis by different
individuals
Pose estimation can help!

18. 18
Aim: automate post-castrated
B6D2F1 mice sexual behaviour

19. Machine predicted mice dyad body part locations
19

20. Machine predicted occurrence ofselected behaviours
20
Classifier colours:
•Mounting or intromission
•Ejaculation

21. Classifier performance depended on parameter settings
21
Parameters 1- 4 = different combos:
•Behaviour probability threshold
•# of subsequent frames to consider
•# of those frames predicted behaviour
???????????????????????????????????????????????????????????????????????? =
???????????????????????????????????????????????? ????????????????????????????????????????????????????????????????????????????????????????????????
???????????????????????????????????????????????? ???????????????????????????????????????????????????????????????????????????????????????????????? +???????????????????????????????????????????????????????????? ????????????????????????????????????????????????????????????????????????????????????????????????
???????????????????????????????????????????????????????????????????????????????????????????????????????????? =
???????????????????????????????????????????????? ????????????????????????????????????????????????????????????????????????????????????????????????
???????????????????????????????????????????????? ???????????????????????????????????????????????????????????????????????????????????????????????? +???????????????????????????????????????????????????????????? ????????????????????????????????????????????????????????????????????????????????????????????????

22. 22
Conclusions & Implications
•Outcome 1: machine predicted mice dyad body parts
•Outcome 2: machine predicted few sexual behaviours
•Significance: neuroscience of steroid-independent rodent sexual behaviour
•Future Application: sexual behaviour health for gender- diverse individuals

23. 23
Dissertation Aim C: Background

24. 24
Some sexual behaviours have a collective context
Adapted from A Rathore, K Isvaran, V Guttal (2022)
Philos Trans R Soc Lond B Biol Sci
T, defending territory
D, displaying
C, courting
F, engaged in fight

25. 25
Constructing the neuroscience of collective behaviour
require studying multiple brains
A Forli & MM Yartsev (2023) Nature

26. 26
Astatotilapia burtoni: lekking species suitable for
studying the collective brain
Adapted from KP Maruska et al. R Wayne (2022) Horm Behav
Dom, dominant
Sub, subordinate

27. 27
Aim: automate collective
behavioural analyses of
East African cichlids

28. 28
Machine predicted body parts of cichlids in communities

29. 29
Assembled body parts to individuals in two separated cichlids setup
Left tank cichlid
Right tank cichlid

30. 30
Kinematics analysed for individual cichlids
0
20
40
60
80
100
120
140
160
180
200
0 2000 4000 6000 8000 10000
Speed
(
pixels
/
second
)
Video frame
dorsal_mid
dorsal_opercular_mid
ventral_opercular
Dorsal_opercular_midDorsal_mid
ventral_opercular
Adapted from K Maruska
et al. C Maximino (2019)
Brain Res

31. 31
Conclusions & Implications
•Outcome 1: machine predicted body parts in cichlids community
•Outcome 2: analysed individual cichlid kinematics in divided setup
•Significance: multi-brain studies in cichlid lekking
•Future Application: improve animal ethics by considering sexual behaviour as welfare

32. 32
Dissertation Aim D: Background

33. 33
Model organisms in the lab have altered brain and behaviour
AJ Turko et al. BG Borowiec (2023) J Exp Biol

34. 34
Sex diversity in the wild are commonly studied with camera traps
Chelonoidischilensis(Ochoa et al., 2021) Odocoileusvirginianus(Peterson et al., 2021) Arctogalidia trivirgata (Barca et al., 2015)
Lophiomys imhausi (Weinstein et al., 2020)Arctogalidia trivirgata (Haysom et al., 2021)Hystrix cristata (Coppola & Felicioli, 2021)Antidorcasmarsupialis(Pardo et al., 2021)
Priodontes maximus(Desbiezet al., 2020)
Aepycerosmelampus(Pardo et al., 2021) Panthera onca (Harmsen et al., 2020)
Budorcas taxicolor bedfordi (Li et al., 2020)Panthera pardus kotiya (Webb et al., 2020) Panthera onca (Eriksson et al., 2022)
ManisPentadactyla(Sun et al., 2021) Panthera pardusmelas(Ario et al., 2022)
Martes caurina(Delheimeret al., 2023)

35. 35
Aim: develop camera trap strategies
for wildlife neuroscience

36. 36
DeepLabCut SuperAnimal: machine prediction of
quadruped body parts
SK Ye et al. MW Mathis (2024) Nat Commun

37. 37
Labelled quadrupeds body parts in camera trap datasets for
fine-tuning DeepLabCut SuperAnimal
A CB
Snapshot Kruger
Wellington Camera Traps
African Field School on Design and Analysis of
Camera Trap Studies 2023

38. 38
Brainstormed some considerations for designing
wildlife neuroscience studies

39. 39
Developed a framework to assess feasibility & ethics of
neuroimaging minimally-captive free -behaving wildlife

40. 40
Conclusion & Implication
•Outcome 1: labelled quadruped body parts in camera trap images
•Outcome 2: developed a framework for wildlife neuroscience
•Significance: neuroimaging wildlife across the tree of life to construct sex diversity
•Future Application: conservation efforts

41. 41
General Conclusion
•Created ejaculate
classifiers for
B6D2F1 hybrid mice
•Engineered cells resembling human sebocytes
•Trained pose estimation model for up to 24 East African cichlids
•Developed strategies for wildlife neuroscience

42. 42
General implications for sex variability and diversity
•Framed sex/gender as non- binary and non-essentialist
•Treated science as diffractive & only one way to construct sex/gender
•Pivoted research to benefit sex/gender individuals

43. 43
Acknowledgements

44. 44
Appendices: General

45. 45
S Spaulding & L Fuselier (2023) Anim Behav

46. 46
AT Zemenicket al. MG Weber (2022) Bioscience

47. 47
KO Smiley et al. SD Sun (2024) Horm Behav

48. 48
KO Smiley et al. SD Sun (2024) Horm Behav

49. 49
K Aghi et al. EM Guthman (2022) Biol Psychiatry Cogn Neurosci Neuroimaging

50. 50
NB Alexander et al. JJ Coon (2024) Wildl Soc Bull

51. 51
JF McLaughlin et al. SE Lipshutz
(2023) Int Comp Biol

52. 52
SM Anders et al. SB Chadwick (2023) Psychol Sex Orientat Gend Divers

53. 53
K Aghi et al. ZY Weinberg (2024) Cell

54. 54
RA Fulgu & V Capraro (2024) PsyArXiv

55. 55
C Fine (2021) Cold Spring Harb Perspect Biol

56. 56
M Dubova, RL Goldstone (2023) Trend Cogn Sci

57. 57
M Dubova, RL Goldstone (2023) Trend Cogn Sci

58. 58
M Dubova, RL Goldstone (2023) Trend Cogn Sci

59. 59
D Rocks, H Cham, M Kundakovic (2022) Biol Sex Diff
This obscuringeffect of “sex” was elegantly
demonstrated in a recent neuroendocrinestudy ,
where researchers analyzed their data by “sex”
and then byestrous stage. Their analysis revealed
that collapsing sex category hidthe dynamic
nature of ovarian hormones (Rocks et al., 2022): if
theexperiment had assumed the internal
coherence of sex categories (as iscommon
practice), important dynamics within these
categories wouldhave been overlooked to the
detriment of our scientific understanding.

60. 60
MBlencowe et al. X Yang (2022) Genome Res
Indeed, ourscientific community has long
recognized that sexed physiologies likegonadal
hormones and sex chromosomes can interact,
synergize, counteract,or simply coexist (Blencowe
et al., 2022) with one another. Thishas been
demonstrated in numerous studies carried out in
the Four CoreGenotypes model, where sex
chromosome complement is divested
fromgonadal development (De Vries et al., 2002)
and has led to discoveriesthat sex chromosomes
interact with hormonal milieu to produce
varyingphenotypes (Chen et al., 2013, 2015),
even “compensating” for eachother to some
extent (Chen et al., 2012).

61. 61
S Naqvi et al. DC Page (2019) Science
In a large study of geneexpression in 12 tissues in
five mammalian species, includinghumans and
mice, Naqvi and colleagues showed that
samplestended to cluster by species, not sex,
suggesting that althoughthe genes themselves
are conserved across species,
sexrelateddifferences in their expression are
not.35 They arguedthat sex bias in the expression
of most genes has evolvedrecently and, as such,
is not shared among most mammals.Even within
species, sex-related variation may not
generalizeto other animals of the same species; in
rodents in particular,strain, parent -offspring
relations, and even litter effects couldbe highly
relevant to the size and direction of any
observedsex-related variation.

62. 62
D Joel, CJ Smith, AH Veenema (2024) Horm Behav
Joel et al. (2024)showed that even for measures that are historically thought to be highlysexually dimorphic in rodents, e.g., measures of
nonapeptide receptordensity, differences are smaller and less persistent than previouslyassumed. In their study, of the 101 brain measures
considered, only onewas differentiated best by sex and reliably dimorphic. For the rest, themagnitude of the sex difference varied according
to factors such as age,social experience, and social context. Like in humans, whether anymeasurement was more “male-like” or “female-
like” was not necessarilywell-predicted by the other measures within individual. That is, individualswith the most male-like value for a
particular measure did notreliably have a male-like value for another.

63. 63

64. 64

65. 65

66. 66

67. 67
V Yadav, S Sun, J Heitman (2023) PNAS

68. 68
AS Dichman (2024) Distinktion
The 2020 art exhibition Heksejagt at the Danish art
center Charlottenborg showcased Rasmus Myrup's
installations, exploring gender and cross-species
boundaries. Myrup created a room of creatures that
blend human and tree characteristics, highlighting
how gender extends beyond humans and intersects
with art, biology, and political theory.These
sculptures, called "wo/od/men," include figures like
Gerd, with a pinecone face and a mix of masculine
and feminine traits, Freya, with a rose face and
feminine attributes, and Skadi, made of pine
needles. The creatures appear to interact,
suggesting a dynamic conversation about gender
identity.Myrup's work encourages us to imagine a
world where gender is fluid and complex. The
wo/od/men embody different levels of human and
plant traits, challenging traditional gender norms.
For example, Freya has a more humanlike gender
identity, while Gerd's identity is more
ambiguous.Tree biology, where some trees are
hermaphroditic or can switch sexes, parallels the
fluidity of gender in Myrup's art. This challenges rigid
human gender norms and illustrates the importance
of interdisciplinary and multispecies approaches to
understanding gender in contemporary politics.

69. 69

70. 70
Appendices: Specific Aim 1

71. 71

72. 72

73. 73

74. 74

75. 75

76. 76Ejaculate classifier

77. 77Mount_intromit classifier

78. 78
JM Siegford et al. GJM Rosa
(2023) App Anim Behav Sci

79. 79
AJ Calhoun & AE Hady (2023)
iScience

80. 80
A Kennedy (2022) Curr Opin Neurobiol

81. 81
A Mathis et al. M Bethge (2018) Nat Neurosci

82. 82
L Kingsbury et al. WZ Hong (2020) Neuron

83. 83
JD Monk et al.CE McDOnough (2019) Nat Ecol Evol

84. 84
C Westlin et al. LF Barrett (2023) Trend Cogn Sci

85. 85
GYR Yang et al. XJ Wang (2020) Neuron

86. 86
YJ Hu et al. B Ye (2022) Cell Rep Methods

87. 87SRO Nilsson et al. SA GOlden (2020) Biorxiv

88. 88S Isik & G Unal (2023) Front Neurosci

89. 89S Isik & G Unal (2023) Front Neurosci

90. 90M Perez & CT Franklin (2023) arXiv

91. 91M Perez & CT Franklin (2023) arXiv

92. 92
reproductive behaviourcontrol column-L Mei, T Osakada, DY Lin (2023) Science

93. 93
LJ Qi et al. DD Ginty(2024) Nature

94. 94
Appendices: Specific Aim 2

95. 95

96. 96

97. 97

98. 98

99. 99

100. 100

101. 101
J Lauer et al. A Mathis (2022) Nat Methods

102. 102
MR Howard et al. B Alward (2024) Horm Behav

103. 103
M Vidal et al. A Mathis (2021) Integr Comp Biol

104. 104
A Rathore, K Isvaran, V Guttal (2022) Philos Trans R Soc Lond B Biol Sci

105. 105
PD Dijkstra et al. TJ Piefke, TR Bonnell (2022) Hydrobiologia

106. 106
JH Yu, JL Napoli, ML Barron (2024) Curr Opin Neurobiol

107. 107
JH Yu, JL Napoli, ML Barron (2024) Curr Opin Neurobiol

108. 108
A Gosztolai & P Ramdya (2022) Curr Opin Neurobiol

109. 109
A Gosztolai & P Ramdya (2022) Curr Opin Neurobiol

110. 110

111. 111
D Aldarondoet al.BP Ölveczky(2024) Nature

112. 112
Appendices: Specific Aim 3

113. 113

114. 114

115. 115
E Edelblutte, R Krithivasan, MN Hayek (2021) Conservation Biol

116. 116
TD Jessenet al. CT Darimont(2021) Front EcolEnv

117. 117
EL Hazen et al.B Abrahms (2024) Annu Rev Environ Resour

118. 118
M Bessonet al. CF Clements (2022) Ecol Lett

119. 119
R Kays & M Wikelski (2023) Trend Ecol Evol

120. 120
M Pichler & F Hartig (2023) Methods Ecol Evol

121. 121

122. 122

123. 123
Appendices: Specific Aim 4

124. 124

125. 125

126. 126

127. 127

128. 128

129. 129

130. 130C Lenschow, ARP Mendes, SQ Lima (2022) Front Neural Circruits

131. 131
PR Pitrez et al. L Ferreira (2024) Nat Commun

132. 132
VEJM Palasantzas et al. JY Fu (2023) Trend Genet

133. 133
RW Claytonet al.R Paus (2020) Biol Rev

134. 134
RW Claytonet al.R Paus (2020) Biol Rev

135. 135
Y Ramo, M Bohm, R Paus (2021) Trend Mol Med

136. 136
VSC Chiang et al. RG Hunter (2022) Front Behav Neurosci

137. 137
LJ Elias et al.IA Saboor (2023) Cell

138. 138
M Pacesa, O Pelea, M Jinek (2024) Cell

139. 139
S Domcke & J Shendure (2023) Cell

140. 140

141. 141
AK Velazquez et al. Z Ignatova (2023) ACS Synth Biol

142. 142
A Azizoglu, R Brent, F Rudolf (2021) Elife

143. 143