Heterochronic shift in gene expression leads to ontogenetic morphological divergence between two closely related polyploid species

Abstract

Heterochrony—alteration to the rate or timing of development—is an important mechanism of trait differentiation associated with speciation. Heterochrony may explain the morphological divergence between two polyploid species, June sucker (Chasmistes liorus) and Utah sucker (Catostomus ardens). The larvae of both species have terminal mouths; however, as adults, June sucker and Utah sucker develop subterminal and ventral mouths, respectively. We document a difference in the timing of shape development and a corresponding change in the timing of gene expression, suggesting the distinctive mouth morphology in June suckers may result from paedomorphosis. Specifically, adult June suckers exhibit an intermediate mouth morphology between the larval (terminal) and ancestral (ventral) states. Endemic and sympatric Chasmistes/Catostomus pairs in two other lakes also are morphologically divergent, but genetically similar. These species pairs could have resulted from the differential expression of genes and corresponding divergence in trait development. Paedomorphosis may lead to adaptive diversification in Catostomids.

Full text

iScience
Article
Heterochronic shift in gene expression leads to
ontogenetic morphological divergence between
two closely related polyploid species
Peter C. Searle,
Dennis K.
Shiozawa, R. Paul
Evans, Jonathon T.
Hill, Arminda Suli,
Michael R. Stark,
Mark C. Belk
pcs222@cornell.edu
Highlights
June and Utah sucker
exhibit synchrony between
gene expression and head
development
Divergence in mouth
morphology is consistent
with paedomorphosis in
June sucker
Incorporating ontogeny
informs evolutionary
change among closely
related species
Searle et al., iScience 27,
109566
April 19, 2024 ª2024 The
Authors. Published by Elsevier
Inc.
https://doi.org/10.1016/
j.isci.2024.109566
ll
OPEN ACCESS

iScience
Article
Heterochronic shift in gene expression leads
to ontogenetic morphological divergence
between two closely related polyploid species
Peter C. Searle,
1,2,5,
*Dennis K. Shiozawa,
2
R. Paul Evans,
3
Jonathon T. Hill,
4
Arminda Suli,
4
Michael R. Stark,
4
and Mark C. Belk
2
SUMMARY
Heterochrony—alteration to the rate or timing of development—is an important mechanism of trait dif-
ferentiation associated with speciation. Heterochrony may explain the morphological divergence be-
tween two polyploid species, June sucker (Chasmistes liorus) and Utah sucker (Catostomus ardens).
The larvae of both species have terminal mouths; however, as adults, June sucker and Utah sucker develop
subterminal and ventral mouths, respectively. We document a difference in the timing of shape develop-
ment and a corresponding change in the timing of gene expression, suggesting the distinctive mouth
morphology in June suckers may result from paedomorphosis. Specifically, adult June suckers exhibit
an intermediate mouth morphology between the larval (terminal) and ancestral (ventral) states. Endemic
and sympatric Chasmistes/Catostomus pairs in two other lakes also are morphologically divergent, but
genetically similar. These species pairs could have resulted from the differential expression of genes
and corresponding divergence in trait development. Paedomorphosis may lead to adaptive diversification
in Catostomids.
INTRODUCTION
Understanding the evolution of biological diversity is a major goal in evolutionary developmental biology because developmental processes
are tightly linked with morphological changes and corresponding evolutionary divergence.
1–3
Gene expression analyses coupled with de-
scriptions of ontogenetic morphological change have recently gained attention as an important method for exploring evolutionary relation-
ships. Changes in developmental gene regulation can lead to substantial morphological divergence as a consequence of altered develop-
mental timing.
4
This process of alteration to the rate or timing of development in a species compared to its ancestor is termed
heterochrony.
5,6
Slight changes to temporal (i.e., the onset or offset of expression, as well as the rate of expression) or spatial gene expression
during periods of rapid ontogenetic change, can shorten or prolong development, significantly altering the appearance of a descendant from
its ancestor.
6,7
Heterochrony has been identified as an important evolutionary mechanism across many taxa.
8–14
Because heterochrony is
often the product of changes in gene regulation, it can result in species that are well differentiated morphologically but show only small ge-
netic differentiation across the genome.
15
Such species are difficult to differentiate when only genetic data are used to determine levels of
differentiation and corresponding phylogenetic relationships.
16
Thus, the explicit incorporation of ontogeny in both developmental morpho-
logical change and in gene expression derived from transcriptome analysis is critical for observing patterns of evolutionary change and diver-
gence among such species.
2,3
Heterochronic shifts in gene expression and corresponding morphological changes may explain the evolutionary divergence between
June sucker (Chasmistes liorus) and its co-occurring sister species, the Utah sucker (Catostomus ardens). June sucker is a highly derived
Lake sucker (genera: Chasmistes,Deltistes, and Xyrauchen, Pisces: Catostomidae) that is endemic to Utah Lake, Utah, USA. June sucker is
adapted for planktivorous feeding and is characterized by a subterminal mouth, lower lip lobes separated by a wide central gap, thin lips
with reduced papillation, and numerous, dendritic gill rakers.
17
Utah sucker is a typical benthic sucker (widely distributed in lakes and rivers
of the intermountain western USA) adapted for the suction feeding and filtration of larger macroinvertebrates from muddy substrates.
18,19
Utah sucker is characterized by a ventral mouth, wide lower lip lobes with no central gap, heavy papillation on the lips, and unbranched
gill rakers.
17
The same pattern of highly derived lake suckers (Chasmistes) co-occurring with closely related benthic suckers (Catostomus)
1
Department of Ecology and Evolutionary Biology, Cornell University, Ithaca, NY 14853, USA
2
Department of Biology, Brigham Young University, Provo, UT 84602, USA
3
Department of Microbiology and Molecular Biology, Brigham Young University, Provo, UT 84602, USA
4
Department of Cell Biology and Physiology, Brigham Young University, Provo, UT 84602, USA
5
Lead contact
*Correspondence: pcs222@cornell.edu
https://doi.org/10.1016/j.isci.2024.109566
iScience 27, 109566, April 19, 2024 ª2024 The Authors. Published by Elsevier Inc.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1
ll
OPEN ACCESS

is repeated multiple times in western North America.
17
In each case, the pattern of differentiation in mouth morphology is the same as
described for the June sucker and the Utah sucker, and the mouth morphology of adults is the most distinctive trait that differentiates the
co-occurring species pairs. In contrast to adults, the morphology of the head and mouth in larvae is extremely similar between lake suckers
and benthic suckers. Larvae of both groups have terminal mouths, relatively large eyes and heads, and relatively small and narrow bodies.
20
Thus, the polarity of development from a similar larval shape to a differentiated adult shape is common to all sympatric lake sucker/benthic
sucker pairs.
These morphological differences between co-occurring sucker species have been well documented in both extant and extinct species.
Taxonomic characterizations based on the shape and position of the mouth formed the basis for the conclusion that Chasmistes and Catos-
tomus represent ancient lineages, and that individuals that exhibited intermediate mouth morphology in Utah Lake were the result of recent
introgression between June sucker and Utah sucker.
21
However, more recent work aimed at using genetic data to definitively differentiate the
species has been unsuccessful at finding consistent patterns of genetic differences. Lack of genetic variation between the two putative species
(June sucker and Utah sucker) in Utah Lake led Mock et al.
19
and Cole et al.
22
to propose that Utah sucker in Utah Lake recently differentiated
into two forms: a planktivorous form (June sucker) and a benthivorous form (Utah sucker; See Smith et al.
17
for a discussion of an alternative
hypothesis why June sucker and Utah sucker lack genetic markers). Thus, the presence of intermediates could represent a convergence to-
ward or divergence away from the June sucker-like morphology, depending on environmental conditions. Cole et al.
22
further suggested that
the sucker complex in Utah Lake could be the result of a long history of reticulated evolution (Reticulated evolution is often defined as gene
transfer between species.
23
Here we define reticulated evolution specifically as described by Cole et al.
22
: ‘‘a long .history of genetically
shallow but morphologically pronounced divergence and convergence, following fluctuating environmental conditions.’’) or could represent
putative lineages that recently diverged (incipient species). However, Cole et al.
22
did not provide any explanations as to how reticulated evo-
lution or recent divergence could occur in these suckers. The condition of little difference in genetic variation but consistent variation in adult
morphology can be further resolved by reference to ontogenetic patterns of development in both gene expression and corresponding
morphology.
In this study, we combine analyses of morphological change and gene expression to reveal differences in the ontogenetic developmental
pattern between June sucker and Utah sucker (Utah sucker is represented by two lineages in our study because it is a widespread species). We
hypothesized that the easily recognizable differences between adults resulted from differential timing, i.e., heterochrony, of development
over the ontogenetic transition from larvae to juveniles. Specifically, we describe and test for synchrony between the expression of genes
associated with the cranial development and development of distinctive head and mouth morphology between the two species. We used
geometric morphometric analyses of head shape, as well as premaxilla shape and size to determine the critical developmental period during
ontogeny when mouth position began to diverge between these species. We then used differential gene expression analyses to determine
general gene expression patterns that coincided with morphological changes, as well as identify specific genes associated with these
morphological changes. We document a difference in the timing of shape development and a corresponding change in the timing of
gene expression between the June sucker and Utah sucker lineages, consistent with a paedomorphic heterochronic change in development
in June sucker. This study provides an important example of the integration of ontogeny into the exploration of differences between closely
related polyploid species.
RESULTS
Head shape undergoes a shorter developmental sequence in June sucker
We used landmark-based geometric morphometrics
24
to determine how head shape changed during the larval to juvenile transition in
one June sucker and two Utah sucker lineages (Utah sucker from Utah Lake, UT, USA and Utah sucker from an introduced population of
fish in Strawberry Reservoir, UT, USA; see STAR Methods for additional discussion). We found that head shape differed significantly
among developmental weeks and among the three lineages, and that week and lineage exhibited a significant interaction (see two-
way and three-way interactions with index variable; Table 1). Relative warp 1, which accounted for 64.59% of the variation in head shape,
captured the shape change that occurred during larval development, including a shift from a larval stage with smaller heads and larger
eyes to a juvenile stage with larger heads and smaller eyes, as well as lineage-specific variation including an elongation of the snout and
more ventral positioning of the mouth in the Utah sucker lineages compared to June sucker (Figure 1A). Initially, on relative warp 1 each
lineage exhibited common shape variation, including rapid shape change from weeks 0–6, followed by a decrease in the magnitude of
shape change from weeks 6–12. After week 12, lineage-specific variation became more apparent, with Utah sucker lineages continuing
to move further to the right along relative warp 1 compared to June sucker (Figure 1B). There was no clear separation among
developmental weeks or lineages along relative warp 2, which accounted for an additional 9.57% of the variation in head shape
(Figure 1B).
Week 14 represents the endpoint of the developmental sequence we explored in this study. At this endpoint, we found that the two
Utah Sucker lineages exhibited similar shapes on all relative warps except relative warp 7. In contrast, shape differences between the
June sucker and one or both Utah sucker lineages occurred on relative warps 1, 2, 4, 5, 6, 7, and 8 (Figure 1C). To visualize total shape
differences among all three lineages, we compared shape differences at week 14 across all relative warps. Consistent with patterns in
Figure 1C, shape was similar between the Utah sucker lineages, but was different between June sucker and each Utah sucker lineage
(Figure 1D). Both Utah sucker lineages exhibited an elongated snout, down-turned mouth, and smaller eye compared to the June sucker
lineage.
ll
OPEN ACCESS
2iScience 27, 109566, April 19, 2024
iScienc
e
Article

Premaxilla shape is different between lineages after 14 weeks of development
The premaxilla is a diagnostic feature used for differentiating extant and extinct species in Catostomidae.
21
Specifically, the premaxilla in
June sucker has a moderately obtuse angle between the premaxillary limbs, and the vertical and lateral limbs are relatively equal; whereas
the premaxilla in Utah sucker has a right angle between the premaxillary limbs, and the vertical limb is longer than the lateral limb (Fig-
ure S1; also, see
21
Figure 10 for examples from adult suckers). We used landmark-based geometric morphometrics
24
of premaxilla shape,
as well as measurements of premaxilla size to determine how shape and size changed during the larval to juvenile transition among these
three lineages.
We found that premaxilla shape differed significantly among developmental weeks and among the three lineages and that both week and
lineage exhibited a significant interaction (see two-way and three-way interactions with index variable; Table 1). Similar to head shape, relative
warp 1 accounted for more than half (54.79%) of the shape variation in our specimens. Relative warp 1 captured the shape change that
occurred during larval development, including a shift from a larval stage, with a relatively longer lateral premaxillary limb in comparison to
the vertical limb, to a juvenile stage with a relative increase in the length of the vertical limb compared to the lateral limb (Figure 2A). On
relative warp 1, each lineage initially underwent rapid shape changes from weeks 3–6, followed by a decrease in the magnitude of shape
change from weeks 6–14 (Figure 2B). Relative warp 2 (22.57%) captured lineage-specific variation, including a slightly less obtuse angle be-
tween the premaxillary limbs in the Utah sucker lineages compared to the June sucker (Figure 2A). Specifically, Utah sucker lineages had
slightly more positive values along relative warp 2 compared to June sucker at week 14 (Figure 2B).
We further explored premaxilla shape differences between lineages at the developmental endpoint represented by week 14. Similar to the
analysis of head shape, we found that the shape of the premaxilla did not differ between the two Utah sucker lineages except on relative warp
6; whereas, the June sucker lineage differed from one or both Utah sucker lineages on relative warps 2, 3, 4, and 6 (Figure 2C). To visualize total
shape differences among all three lineages, we compared shape differences at week 14 across all relative warps. Consistent with the analysis
of head shape, we found that shape was different between the June sucker and both Utah sucker lineages (Figure 2D) at week 14. June suckers
show a more obtuse angle between the two premaxillary limbs compared to both of the Utah sucker lineages. However, the comparison be-
tween the two Utah sucker lineages revealed that the shape of the premaxilla of the Utah Lake Utah sucker was intermediate between the June
sucker and the Strawberry Reservoir Utah sucker.
Premaxilla size undergoes a shorter developmental sequence in the June sucker
We used linear measurements of the premaxilla to determine how the size of the premaxilla changed during the larval to juvenile transition. A
plot of premaxilla size (defined as the sum of the length of the vertical and lateral limbs) revealed that growth rate and corresponding size were
similar among all three lineages at 4, 6, 8, and 10 weeks. Growth of the premaxilla accelerated in the two Utah sucker lineages but slowed in
the June sucker lineage from weeks 10–14 (Figure 3), consistent with changes in head shape among the three lineages. Unlike the head shape,
but consistent with the shape of the premaxilla, the final size of the premaxilla in Utah Lake Utah sucker was intermediate between June sucker
and Strawberry Reservoir Utah sucker.
De novo transcriptomes are comparable between lineages
We de novo assembled transcriptomes using PacBio sequences (560 gigabytes of sequence data with 5,691,721 reads) for each suckerlineage
resulting in 36,866; 35,656; and 32,115 transcripts for June sucker, Utah Lake Utah sucker, and Strawberry Reservoir Utah sucker, respectively
Table 1. Multivariate analysis of variance effects for head shape and premaxilla shape of June sucker and two lineages of Utah sucker
Source Degrees of Freedom F-Value p Value
Head Shape
Lineages (Ln) 2, 1809 2.4 0.0910
Week 14, 1809 30.23 <0.0001
Index 8, 1471 7.15 <0.0001
Ln*Index 16, 1982 16.27 <0.0001
Week*Index 112, 3079 13.01 <0.0001
Ln*Week*Index 243, 3198 1.72 <0.0001
Premaxilla Shape
Lineages (Ln) 2, 558 2.69 0.0687
Week 5, 558 4.38 0.0006
Index 5, 510 0.29 0.9178
Ln*Index 10, 653 6.37 <0.0001
Week*Index 25, 831 7.53 <0.0001
Ln*Week*Index 60, 930 1.68 0.0012
ll
OPEN ACCESS
iScience 27, 109566, April 19, 2024 3
iScienc
e
Article

(Table S1). We validated our de novo transcriptomes by running BUSCO (v5.1.2)
25
and performing a functional annotation. BUSCO revealed
that the transcriptomes contained 77.5% complete BUSCOs (highly conserved single-copy orthologs) for all lineages, with a high number of
duplicate BUSCOs (Figure S2). The high percentage of complete, duplicated BUSCOs is consistent with the tetraploid origin of suckers. In
addition, the transcriptomes had 27.5% missing BUSCOs. This is likely because we only sequenced the head of the lineages and would
expect transcripts from other tissues to be missing from the transcriptomes.
01
23 45 6 7 8
91011
12
13
14
1 0 2 3 4 5 7 6 8
9 10
1112 13 14
0234 5 6
7 89
10 11
12 13 14
June Utah Strawberry
Relative Warp 1 (64.59%)
Relative Warp 2 (9.57%)
0.00
0.04
0.08
0.12
123456789
Relative Warp Number
Estimate
June
Utah
Strawberry
Relative Warp 1 (-)
Relative Warp 2 (-)
Relative Warp 1 (+)
Relative Warp 2 (+)
June Sucker
Utah Sucker (UL) Utah Sucker (SR)
A B
C D
Figure 1. Head shape undergoes a shorter developmental sequence in June sucker
Results of geometric morphometric analysis of head shape.
(A) Thin plate spline deformations from positive and negative extremes of relative warps 1 and 2.
(B) Least squares means of head shape by week for June sucker, Utah Lake Utah sucker, and Strawberry Reservoir Utah sucker on relative warps 1 and 2.
(C) Least squares means (error bars represent 95% confidence intervals of the mean) for each sucker lineage on the 9 relative warps from the multivariate linear
mixed model of head shape variation at week 14.
(D) Thin plate spline deformations of divergence in head shape between sucker lineages at week 14. UL = Utah Lake, SR = Strawberry Reservoir.
ll
OPEN ACCESS
4iScience 27, 109566, April 19, 2024
iScienc
e
Article

The functional annotation showed that the de novo transcriptomes were also similar across several metrics. A total of 31,538 (85.55%),
29,829 (83.66%), and 27,250 (84.85%) transcripts were annotated for June sucker, Utah Lake Utah sucker, and Strawberry Reservoir Utah sucker,
respectively. Each transcriptome shared blast hits against the same top ten species, all from the order Cyrpiniformes (June sucker, 91.78%;
Utah Lake Utah sucker, 90.89%; Strawberry Reservoir Utah sucker, 91.91%). The three top gene ontology categories (Level 2) for all lineages
46810
12
14
46 8
10
12 14
4
6
8
10
12
14
June Utah Strawberry
−0.04
00.000.0
0.04
−0.04
00.000.0
0.04
−0.04
00.000.0
0.04
0.0
Relative Warp 1 (54.79%)
Relative Warp 2 (22.57%)
Relative Warp 1 (-)
Relative Warp 2 (-)
Relative Warp 1 (+)
Relative Warp 2 (+)
−0.04
−0.02
0.00
0.02
0.04
0.06
123456
Relative Warp Number
Estimate
June
Utah
Strawberry
June Sucker
Utah Sucker (UL) Utah Sucker (SR)
A B
C D
Figure 2. Premaxilla shape is different between lineages after 14 weeks of development
Results of geometric morphometric analysis of premaxilla shape.
(A) Thin plate splines deformations from positive and negative extremes of relative warps 1 and 2.
(B) Least squares means of premaxilla shape by week for June sucker, Utah Lake Utah sucker and Strawberry Reservoir Utah sucker on relative warps 1 and 2.
(C) Least squares means (error bars represent 95% confidence intervals of the mean) for each sucker lineage on the 6 relative warps from the multivariate linear
mixed model of premaxilla shape variation at week 14.
(D) Thin plate spline deformations of divergence in premaxilla shape between lineages at week 14. UL = Utah Lake, SR = Strawberry Reservoir.
ll
OPEN ACCESS
iScience 27, 109566, April 19, 2024 5
iScienc
e
Article

were cellular anatomical entity (cellular component), cellular process (biological processes), and binding (molecular function). Transferases
were the prominent enzyme category, with the top subclass being the phospho-transferase group for all lineages. Taken together, these an-
alyses indicated that we had high-quality transcriptomes that were comparable between lineages.
General and specific gene expression patterns are consistent with shorter development in June sucker
We analyzed tissue-specific transcript expression patterns during periods of common (9–12) and lineage-specific (12–14) shape (head) and
size (premaxilla) variation. We used tissue at the tip of the rostrum, directly in front of the eye, to avoid identifying differentially expressed
transcripts associated with the development of other anatomical structures. Multi-species comparisons of gene expression are inherently
challenging because of differences in transcript sequence and transcript length.
26
One approach to explore transcript expression across spe-
cies is to generate lists of differentially expressed transcripts (DETs) within a species and then compare these lists between species.
27,28
We
used this method to explore general transcript expression patterns in these lineages for each pairwise comparison of time points.
During periods of common shape and size variation (pairwise comparison of weeks 9 and 12), relatively few transcripts were differentially
expressed in each lineage. Specifically, the Strawberry Reservoir Utah sucker had 5 DETs, June sucker had 18 DETs, and Utah Lake Utah sucker
had 356 DETs. During weeks 9 and 12, only one DET was shared between the Utah sucker lineages. None of the remaining DETs were shared
between lineages. During periods of lineage-specific shape and size variation (pairwise comparisons of weeks 9 and 14, as well as 12 and 14),
June sucker had the fewest DETs (7 and 29), followed by Utah Lake Utah sucker (960 and 1030), and Strawberry Reservoir Utah sucker with the
most DETs (2757 and 3364). During weeks 9 to 14 and 12 to 14, the Utah sucker lineages shared 212 and 260 common DETs, respectively. No
DETs were shared across all lineages (see Figure 4 for comparisons between other lineages).
Of the 50 candidate transcripts (34 genes) identified—with known association with craniofacial development—the largest classification of
transcripts was those that were unchanged during weeks 9 and 12 in all lineages and only differentially expressed in Strawberry Reservoir Utah
sucker during weeks 12 and 14 (36 transcripts: 10 upregulated, 26 downregulated). The second largest classification of transcripts was those
that were unchanged during weeks 9 and 12 in all lineages and only differentially expressed in Utah Lake Utah sucker during weeks 12 and 14
(9 transcripts: 6 upregulated, 3 downregulated). In addition, none of these 50 candidate transcripts were differentially expressed in the June
sucker during any of the weeks (Table 2;Figures S7 and S8).
1000
1500
2000
468101214
Total Length (μm)
June
Utah
Strawberry
Figure 3. Premaxilla size undergoes a shorter developmental sequence in June sucker
Plot of mean total limb length of the premaxilla (error bars represent 95% confidence intervals of the mean).
ll
OPEN ACCESS
6iScience 27, 109566, April 19, 2024
iScienc
e
Article

DISCUSSION
June sucker are paedomorphic
We aimed to explore how morphology and gene expression patterns change during the larval to juvenile transition in June sucker and two
Utah sucker lineages. Using geometric morphometrics and RNA-seq analyses, we identified a differencein the timing of shape development
and a corresponding change in the timing of gene expression. We add to the growing number of studies that have successfully used RNA-seq
to unravel the complicated relationship between morphology and gene expression.
29–31
Specifically, our results suggest that the morphology
in the June sucker may be the result of paedomorphosis, in which June sucker exhibits a less developed morphology compared to the ances-
tral adult with an intermediate mouth between that of the larval (terminal) and ancestral (ventral) states.
Our analyses of head shape and premaxilla size revealed that the restructuringof mouth position took less time to develop in the June sucker
lineage compared to Utah sucker lineages. The June sucker’s mouth reached a subterminalposition and stopped changing around week 10–12,
while the Utah sucker lineages’ mouths continued to shift ventrally throughweek 14 (Figures 1 and 3). We used premaxilla shape to analyze how
mouth shape changed between lineages. We found that although premaxilla shape began to diverge among lineages around week 14, these
differences were subtle (Figure 2). During morphogenesis, different anatomical structures develop at different rates. Thus, mouth shape may
take longer to develop in these species compared to mouth position. Alternatively, we were limited in the number of specimens we could stain
and clear for the analysis of the premaxilla shape. Increasing our sample size for each week may have revealed a clearer pattern.
Our RNA-seq analyses were consistent with our geometric morphometric analyses. Each lineage experienced common expression pat-
terns during weeks 9 to 12—a period of common shape variation—with few DETs in each lineage. During weeks 12 to 14, lineage-specific
18
355
4
0
0
1
0
4
746
2544
2
1
212
0
12
768
3089
2
15
260
0
June
Utah
Strawberry June Strawberry June
Utah
Strawberry
9v12 9v14 12v14
Utah
Figure 4. General transcript expression patterns are consistent with shorter development in June sucker
Venn diagrams of differentially expressed transcripts for all pairwise comparisons (9v12, 9v14, 12v14) for each lineage. Darker colors indicate a higher numbers of
differentially expressed transcripts.
Table 2. Classifications of 50 candidate transcripts for contrasts from weeks 9 to 12 and 12 to 14
Week 9 v 12 Week 12 v 14
Total TranscriptsJS UL SR JS UL SR
00 0 00 – 26 aldh1a2_1, aldh1a2_2, cldn5_1, crebbp_1, crebbp_2,
crebbp_3, fgfr2_1, foxn3_1, foxn3_4, insig2_2, kat6a_1,
map2k2_1, map2k2_2, nipbla_1, pax9_1, pax9_4,
pitx2_2, plekha1_1, plekha1_3, rab23_1, rras_1,
rras_2, ski_1, ski_2, tgfb1_1, wnt11_1
00 0 00 +10 ankrd11_2, bbs4_1, cfdp1_1, chd7_2, map2k1_2,
plekha1_2, sox3_1, sox3_2, tiparp_2, tiparp_3
00 0 0+ 0 6 mmp14_1, mmp2_1, mmp2_2, schip1_1,
tgfb3_1, tgfb3_2
00 0 0– 0 3 grhl3_1, grhl3_2, sgpl1_1
00 0 0+ – 2 mmp14_2,mmp16_1
0+ 0 0– 0 2 insig1_1,mapk3_1
00 0 0– +1 tgfbr2_1
Unchanged (0), Upregulated (+), and Downregulated (). JS = June sucker, UL = Utah Lake Utah sucker, and SR = Strawberry Reservoir Utah sucker. Underscore
followed by a number indicates which cluster a transcript came from when there were multiple transcripts annotated as the same gene.
ll
OPEN ACCESS
iScience 27, 109566, April 19, 2024 7
iScienc
e
Article

global expression patterns became apparent between the lineages, with the number of DETs increasing dramatically in the Utah sucker lin-
eages compared to the June sucker (Figure 4). These suites of genes that were being upregulated and downregulated were associated with
the active anatomical restructuring in the Utah sucker lineages as the mouth continued to shift into a ventral position in the Utah sucker lin-
eages. However, June sucker maintained the same expression patterns seen from weeks 9 to 12 (i.e., very few DETs), consistent with the short-
ened developmental sequence in these fish.
Utah sucker lineages exhibit differences during the development
Despite the Utah sucker lineages being more like each other than to the June sucker lineages, there were still noticeable differences between
these lineages in our analyses. Specifically, we found that (1) Utah Lake Utah sucker premaxilla shape and size were intermediate between
June sucker and Strawberry Reservoir Utah sucker at week 14 (Figures 2D and 3), (2) the total number of DETs was variable between these
lineages, and (3) the number of shared DETs between each lineage was small compared to the total number of DETs in each lineage at weeks
12 to 14 (Figure 4). The differences between transcript expression in the Utah sucker lineages became more apparent when we analyzed the
expression of individual transcripts. Of the 50 candidate transcripts we identified, the majority were only differentially expressed in Strawberry
Reservoir Utah sucker during weeks 12 to 14 (36 transcripts), followed by transcripts that were only differentially expressed in Utah Lake Utah
sucker during weeks 12 to 14 (9 transcripts). In fact, none of these transcripts shared common expression patterns in the Utah sucker lineages
(Table 2;Figures S7 and S8).
Several potential genetic, ecological, and morphological differences between these lineages could account for this variation. Although
Utah suckers in Strawberry Reservoir were introduced, we do not know the source population of these fish. Thus, we do not know how
long this population has been separated from Utah Lake Utah sucker and what population-level genetic variation may exist between these
lineages. In addition, these lineages were collected from systems that have different abiotic factors. Strawberry Reservoir is a man-made
waterbody at an elevation of 2320 m, with an average depth of 14 m and a maximum depth of 70 m,
32
while Utah Lake is a natural lake at
an elevation of 1368 m, with an average depth of 2.4 m and a maximum depth of 3.7 m
33
The variation between these lineages could be indic-
ative of local adaptation to different growing seasons. Finally, the Utah Lake Utah sucker brood lot has been in captivity for some time (these
fish represent at most an F2 generation as these fish are long-lived and fish were replaced using within-lot crosses), while Strawberry Reservoir
Utah sucker was spawned at Strawberry Reservoir. Thus, the differences between the Utah sucker lineages could be the result of a plastic
response in morphology to the captive environment.
34
Differences between Utah sucker lineages could also be because Catostomids are allopolyploids.
35
Polyploidy is known to negatively
impact the quality and accuracy of gene expression analyses.
36
This can be due to the difficulty of clustering assembled transcripts into pu-
tative genes,
36
as well as functional divergence and expression differences between paralogous genes.
37
In our transcriptomes, we often
identified multiple transcripts annotated as the same gene in each lineage. Clustering these transcripts, revealed that transcripts within clus-
ters were highly similar (comparing between lineages), but transcripts between these clusters (comparing within lineages) were highly dissim-
ilar (see Table S3 for example from mmp2). These groups of homologous sequences could represent isoforms, alleles, orthologs, and/or pa-
ralogs. In addition, we found evidence of conserved (aldh1a2,crebbp,foxn3,grhl3,map2k2,mmp2,pax9,rras,ski,sox3,tgfb3,tiparp) and
diverged (mmp14,plekha1) expression patterns between these homologous sequences (Figures S7 and S8). A recent assembly of the
genome of the Chinese sucker (Myxocyprinus asiaticus), identified high synteny between homeologous chromosomes, suggesting that in
Catostomids, both subgenomes retain some level of function.
35
The presence of many homologous sequences experiencing similar expres-
sion patterns is consistent with the presence of functionally redundant paralogs in these species. Additional analyses would be useful in order
to identify which genes are experiencing relaxed selection, resulting in functionally redundant paralogs, and which genes are undergoing
strong positive selection, resulting in more expression divergence.
37
Any number of these issues could have contributed to the variation
seen between the Utah sucker lineages.
Mouth development genes are only differentially expressed in Utah sucker lineages
Although the 50 candidate transcripts we identified did not share common expression patterns between the Utah sucker lineages, none of
these transcripts were differentially expressed in the June sucker. Because these transcripts are known to be associated with craniofacial
development, these findings are consistent with a lack of anatomical restructuring in June sucker resulting from a shortened developmental
sequence.
For example, mmp2 is known to play roles in craniofacial skeketal morphogenesis, specifically the degradation of the extracellular ma-
trix.
38–40
Ahi et al.
41
identified mmp2 as an important gene associated with the development of Arctic char morphotypes (Salvelinus)in
Lake Thingvallavatn, Iceland. Arctic char in this system has been differentiated into four phenotypic morphs, including a large benthivore,
small benthivore, pelagic planktivore, and pelagic piscivore. Ahi et al.
41
found that the expression of mmp2 was higher in the head of the
benthic morphotypes compared to the pelagic morphotypes across all embryonic stages tested. Our plots of normalized counts for both
transcripts annotated as mmp2 show that expression levels of mmp2 are similar between all lineages from weeks 9 to 12. However, during
week 14, mmp2 expression in the Strawberry Reservoir Utah sucker (benthic morph) was slightly elevated above that in the June sucker
(pelagic morph; Figure 5).
Nipbl—in conjunction with the cohesin protein complex—is know to modify the expression of key developmental regulatory genes.
42
Inac-
tivation of Nipbl in mice and nipbla in zebrafish leads to craniofacial dysmorphia, consistent with Cornelia De Lange Syndrome (down-turned
lips, upturned nose, and compromised bone and cartilage development).
43,44
We found that nipbla (initially identified by eggNOG-mapper
ll
OPEN ACCESS
8iScience 27, 109566, April 19, 2024
iScienc
e
Article

as nipbl, but later confirmed to be nipbla using BLAST,
45
see ‘‘Limitations of the study’’) expression was similar between all lineages from
weeks 9 to 12. However, during week 14, nipbla expression in Strawberry Reservoir Utah sucker was downregulated below the June sucker
and Utah Lake Utah sucker (Figure 5). Although the differences in expression level are small between these lineages, development has been
shown to be extremely sensitive to slight changes in Nipbl activity.
42
Both mmp2 and nipbla show promising expression patterns that need to
be analyzed further.
June sucker and Utah sucker are in the ‘‘messy middle’’ of the speciation continuum
Speciation is best viewed as a continuous process (speciation continuum), rather than a discrete event.
46
However, drawing a line on this con-
tinuum often leads to conflicting results, especially in closely related, recently diverged species. It is possible that the June sucker and Utah
sucker are reticulating or are incipient species in the ‘‘messy middle’’ (sensu
47
) of the speciation continuum. How speciation and ontogenetic
morphological divergence proceeds in Utah Lake, may be influenced by long-term variation in water levels. Since its formation about 10,000
years ago as a remnant of Lake Bonneville,
48
Utah Lake has undergone subsequent expansions and contractions. Evidence suggests that dur-
ing periods of drought, Utah Lake may have become significantly dessicated.
22
This pattern of alternating environmental conditions may
contribute to the variation seen in the June sucker and Utah sucker along the benthic-limnnetic axis. During wet periods, increased depth
would favor limnetic morphs due to the increased relative availability of zooplankton. During droughts, benthic feeding and associated
benthic morphology would be favored. The fluctuating conditions may lead to the persistence of intermediates and a prolonged period
of incomplete separation between the two species. Because of this incomplete separation, different lines of evidence typically used to define
species produce conflicting results.
49
Although there is no universally accepted species concept, there is agreement that species delimitation
should include multiple lines of evidence, such as morphology, ecology, and genetic patterns, considered in an ontogenetic developmental
framework.
2,3
Our morphometric and gene expression analyses provide parallel lines of evidence that the June sucker and Utah sucker are
genetically shallow, yet morphologically diverged species in Utah Lake.
Broader Implications: Paedomorphosis in other Chasmistes/Catostomus pairs
June sucker and Utah sucker are one of the four Chasmistes and Catostomus pairs whose taxonomic relationships have been challenging to
discern. These pairs exhibit similar morphological (subterminal vs. ventral mouths) and genetic patterns (lack of genetic markers that consis-
tently separate these species) as seen in the June sucker and Utah sucker. In addition, in all but Chasmistes cujus and Catostomus funeiventris,
these species pairs are sympatric and fish exist with intermediate mouth morphology.
17,22,50,51
Paedomorphosis—resulting from the differ-
ential expression of genes in response to environmental cues—could be the mechanism that drove the convergent evolution of these Chas-
mistes/Catostomus pairs. These findings are consistent with studies across many taxa,
52–54
particularly amphibians,
55
which show that differ-
ential gene expression leads to paedomorphisis. Studies of similar designs will be required to determine if paedomoprhosis is general among
these Chasmistes/Catostomus pairs.
*
**
10
100
1000
10000
100000
91214
mmp2_1
*
**
10
100
1000
10000
100000
91214
mmp2_2
****
****
10
100
1000
91214
nipbl_1
Species June Utah Strawberry
Time
Counts
Figure 5. Mouth development genes are only differentially expressed in Utah sucker lineages
Plots of the transcript-level expression of mmp2 (represented by two transcripts) and nipbla.
ll
OPEN ACCESS
iScience 27, 109566, April 19, 2024 9
iScienc
e
Article

Limitations of the study
Proper annotation of transcriptomes in polyploid species can be difficult. Transcripts annotated as the same gene—within and between spe-
cies—can be isoforms, alleles, orthologs, and/or paralogs. Thus, it is critical to carefully determine the correct relationships between these
transcripts for gene expression analyses. In addition, eggNOG-mapper sometimes predicted gene names, that were general, and not specific
to fish. For example, eggNOG-mapper identified several transcripts as nipbl although two nipbl genes are present in zebrafish (nipbla or
nipblb). Well-annotated, reference genomes from Chasmistes and Catostomus would be ideal for future analyses of gene expression in
Catostomids.
STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:
dKEY RESOURCES TABLE
dRESOURCE AVAILABILITY
BLead contact
BMaterials availability
BData and code availability
dEXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS
BJune sucker and Utah sucker
dMETHOD DETAILS
BGeometric morphometrics
BSequencing
BTranscriptome assembly
BFunctional annotation
BDifferential expression
BCandidate genes
dQUANTIFICATION AND STATISTICAL ANALYSIS
BGeometric morphometrics
BDifferential expression
SUPPLEMENTAL INFORMATION
Supplemental information can be found online at https://doi.org/10.1016/j.isci.2024.109566.
ACKNOWLEDGMENTS
This study was supported through grants by P.C.S. from the Utah Chapter of the American Fisheries Society, Monte L. Bean Life Science
Museum, Charles Redd Center for Western Studies, and Desert Fishes Council; D.K.S. from the Department of Biology, Brigham Young Uni-
versity; R.P.E. from the Department of Microbiology and Molecular Biology, Brigham Young University; and M.C.B. from the Department of
Biology, Brigham Young University.
We thank the Utah Division of Wildlife Resources Fisheries Experiment Station for rearing and raising the sucker lineages. We thank Eric
Bingham for collecting and preserving samples each week. We thank Jillian Campbell and Spencer Ellsworth for landmarking the photos. We
thank Andrea Kokkonen for helping accession the samples into the Monte L. Bean Life Science Museum. We thank the Office of Research
Computing at Brigham Young University for providing supercomputing resources.
AUTHOR CONTRIBUTIONS
Conceptualization: D.K.S., R.P.E., and M.C.B; methodology: P.C.S., D.K.S., R.P.E., J.T.H., A.S., M.R.S., and M.C.B; validation: P.C.S.; formal
analysis: P.C.S.; investigation: P.C.S.; resources: D.K.S., R.P.E., J.T.H., A.S., M.R.S., and M.C.B; data curation: P.C.S.; writing – original draft:
P.C.S.; writing – reviewing and editing: P.C.S., D.K.S., R.P.E., J.T.H., A.S., M.R.S., and M.C.B; visualization: P.C.S.; supervision: M.C.B.; project
administration: M.C.B.; funding acquisition: P.C.S., D.K.S., R.P.E., and M.C.B.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: May 24, 2023
Revised: November 4, 2023
Accepted: March 22, 2024
Published: March 25, 2024
ll
OPEN ACCESS
10 iScience 27, 109566, April 19, 2024
iScienc
e
Article

REFERENCES
1. Sku
´lason, S., Parsons, K.J., Svanba
¨ck, R.,
Ra
¨sa
¨nen, K., Ferguson, M.M., Adams, C.E.,
Amundsen, P.A., Bartels, P., Bean, C.W.,
Boughman, J.W., et al. (2019). A way forward
with eco evo devo: an extended theory of
resource polymorphism with postglacial
fishes as model systems. Biol. Rev. 94, 1786–
1808. https://doi.org/10.1111/brv.12534.
2. Martynov, A., and Korshunova, T. (2022).
Multilevel Organismal Diversity in an
Ontogenetic Framework as a Solution for the
Species Concept. In Cryptic Species:
Morphological Stasis, Circumscription, and
Hidden Diversity, A.K. Monro and S.J. Mayo,
eds. (Cambridge University Press),
pp. 78–129. https://doi.org/10.1017/
9781009070553.004.
3. Martynov, A., Lundin, K., and Korshunova, T.
(2022). Ontogeny, Phylotypic Periods,
Paedomorphosis, and Ontogenetic
Systematics. Front. Ecol. Evol. 10, 806414.
https://doi.org/10.3389/fevo.2022.806414.
4. Prud’homme, B., Gompel, N., and Carroll,
S.B. (2007). Emerging principles of regulatory
evolution. Proc. Natl. Acad. Sci. USA 104,
8605–8612. https://doi.org/10.1073/pnas.
0700488104.
5. Smith, K.K. (2003). Time’s arrow:
heterochrony and the evolution of
development. Int. J. Dev. Biol. 47, 613–621.
6. McNamara, K.J. (2012). Heterochrony: the
evolution of development. Evol. Edu.
Outreach 5, 203–218. https://doi.org/10.
1007/s12052-012-0420-3.
7. Tautz, D. (2000). Evolution of transcriptional
regulation. Curr. Opin. Genet. Dev. 10,
575–579. https://doi.org/10.1016/s0959-
437x(00)00130-1.
8. Mesa, M.L., Arneri, E., Caputo, V., and
Iglesias, M. (2005). The transparent goby,
Aphia Minuta, review of biology and fisheries
of a paedomorphic european fish. Rev. Fish
Biol. Fish. 15, 89–109. https://doi.org/10.
1007/s11160-005-1613-4.
9. C¸ic¸ek, K., and Ayaz, D. (2011). New data on
facultative paedomorphism of the smooth
newt, Lissotriton vulgaris, in Western
Anatolia, Turkey. J. Freshw. Ecol. 26, 99–103.
https://doi.org/10.1080/02705060.2011.
553924.
10. Bhullar, B.-A.S., Maruga
´n-Lobo
´n, J., Racimo,
F., Bever, G.S., Rowe, T.B., Norell, M.A., and
Abzhanov, A. (2012). Birds have
paedomorphic dinosaur skulls. Nature 487,
223–226. https://doi.org/10.1038/
nature11146.
11. Waller, B.M., Peirce, K., Caeiro, C.C.,
Scheider, L., Burrows, A.M., McCune, S., and
Kaminski, J. (2013). Paedomorphic facial
expressions give dogs a selective advantage.
PLoS One 8, e82686. https://doi.org/10.1371/
journal.pone.0082686.
12. McMahon, D.P., and Hayward, A. (2016). Why
grow up? A perspective on insect strategies
to avoid metamorphosis. Ecol. Entomol. 41,
505–515. https://doi.org/10.1111/een.12313.
13. Sto
¨hr, S., and Martynov, A. (2016).
Paedomorphosis as an evolutionary driving
force: Insights from deep-sea brittle stars.
PLoS One 11, e0164562. https://doi.org/10.
1371/journal.pone.0164562.
14. Scherz, M.D., Hutter, C.R., Rakotoarison, A.,
Riemann, J.C., Ro
¨del, M.O., Ndriantsoa, S.H.,
Glos, J., Hyde Roberts, S., Crottini, A.,
Vences, M., and Glaw, F. (2019).
Morphological and ecological convergence
at the lower size limit for vertebrates
highlighted by five new miniaturised
microhylid frog species from three different
Madagascan genera. PLoS One 14,
e0213314. https://doi.org/10.1371/journal.
pone.0213314.
15. West-Eberhard, M.J. (2003). Developmental
Plasticity and Evolution (Oxford University
Press).
16. Mason, N.A., and Taylor, S.A. (2015).
Differentially expressed genes match bill
morphology and plumage despite largely
undifferentiated genomes in a Holarctic
songbird. Mol. Ecol. 24, 3009–3025. https://
doi.org/10.1111/mec.13140.
17. Smith, G.R., Zaroban, D.W., High, B., Sigler,
J.W., Schilling, J., Krabbenhoft, T.J., and
Dowling, T.E. (2018). Fishes of the Mio-
Pliocene Western Snake River Plain and
vicinity III. Introgressive mtDNA transfer in
hybrid lake suckers (Teleostei, Catostomidae)
in Western United States. Misc. Publ. Mus.
Zool. Univ. Mich. 204, 87–117.
18. Searle, P.C., Linde, J.B., Campbell, J.R.,
Kokkonen, A.L., Shiozawa, D.K., Belk, M.C.,
and Evans, R.P. (2022). Complete
mitochondrial genomes of June sucker and
Utah sucker (Chasmistes liorus and
Catostomus ardens). Mitochondrial DNA Part
B7, 560–562. https://doi.org/10.1080/
23802359.2022.2055984.
19. Mock, K.E., Evans, R.P., Crawford, M., Cardall,
B.L., Janecke, S.U., and Miller, M.P. (2006).
Rangewide molecular structuring in the Utah
sucker (Catostomus ardens). Mol. Ecol. 15,
2223–2238. https://doi.org/10.1111/j.1365-
294X.2006.02932.x.
20. Snyder, D.E., and Muth, R.T. (1988).
Description and identification of June, Utah,
and mountain sucker larvae and early
juveniles. Utah State Div. Wildl. Resour.
21. Miller, R.R., and Smith, G.R. (1981).
Distribution and evolution of Chasmistes
(Pisces: Catostomidae) in western North
America. Occas. Pap. Mus. Zool. Univ. Mich.
696, 1–46.
22. Cole, D.D., Mock, K.E., Cardall, B.L., and
Crowl, T.A. (2008). Morphological and
genetic structuring in the Utah Lake sucker
complex. Mol. Ecol. 17, 5189–5204. https://
doi.org/10.1111/j.1365-294X.2008.03990.x.
23. C¸ oraman, E., Dietz, C., Hempel, E.,
Ghazaryan, A., Levin, E., Presetnik, P.,
Zagmajster, M., and Mayer, F. (2019).
Reticulate evolutionary history of a Western
Palaearctic Bat Complex explained by
multiple mtDNA introgressions in secondary
contacts. J. Biogeogr. 46, 343–354. https://
doi.org/10.1111/jbi.13509.
24. Bookstein, F.L. (1991). Morphometric Tools
for Landmark Data: Geometry and Biology
(Cambridge University Press).
25. Seppey, M., Manni, M., and Zdobnov, E.M.
(2019). BUSCO: Assessing genome assembly
and annotation completeness. Methods Mol.
Biol. 1962, 227–245. https://doi.org/10.1007/
978-1-4939-9173-0_14.
26. Munro, C., Zapata, F., Howison, M., Siebert,
S., and Dunn, C.W. (2022). Evolution of Gene
Expression across Species and Specialized
Zooids in Siphonophora. Mol. Biol. Evol. 39,
msac027. https://doi.org/10.1093/molbev/
msac027.
27. Han, E.-S., and Hickey, M. (2005). Microarray
evaluation of dietary restriction. J. Nutr. 135,
1343–1346. https://doi.org/10.1093/jn/135.
6.1343.
28. von Wyschetzki, K., Rueppell, O., Oettler, J.,
and Heinze, J. (2015). Transcriptomic
signatures mirror the lack of the fecundity/
longevity trade-off in ant queens. Mol. Biol.
Evol. 32, 3173–3185. https://doi.org/10.1093/
molbev/msv186.
29. Albertson, R.C., Yan, Y.L., Titus, T.A., Pisano,
E., Vacchi, M., Yelick, P.C., Detrich, H.W., and
Postlethwait, J.H. (2010). Molecular
pedomorphism underlies craniofacial skeletal
evolution in Antarctic notothenioid fishes.
BMC Evol. Biol. 10,4.https://doi.org/10.
1186/1471-2148-10-4.
30. Gunter, H.M., Koppermann, C., and Meyer, A.
(2014). Revisiting de Beer’s textbook example
of heterochrony and jaw elongation in fish:
Calmodulin expression reflects heterochronic
growth, and underlies morphological
innovation in the jaws of belonoid fishes.
EvoDevo 5,8.https://doi.org/10.1186/2041-
9139-5-8.
31. Powder, K.E., Milch, K., Asselin, G., and
Albertson, R.C. (2015). Constraint and
diversification of developmental trajectories
in cichlid facial morphologies. EvoDevo 6, 25.
https://doi.org/10.1186/s13227-015-0020-8.
32. Baldwin, C.M., Beauchamp, D.A., and Van
Tassell, J.J. (2000). Bioenergetic assessment
of temporal food supply and consumption
demand by Salmonids in the Strawberry
Reservoir food web. Trans. Am. Fish. Soc. 129,
429–450. https://doi.org/10.1577/1548-
8659(2000)129<0429:BAOTFS>2.0.CO;2.
33. Shiozawa, D.K., and Barnes, J.R. (1977). The
microdistribution and population trends of
larval Tanypus stellatus Coquillett and
Chironomus frommeri Atchley and Martin
(Diptera: Chironomidae) in Utah Lake, Utah.
Ecology 58, 610–618. https://doi.org/10.
2307/1939010.
34. Belk, M.C., Benson, L.J., Rasmussen, J., and
Peck, S.L. (2008). Hatchery-induced
morphological variation in an endangered
fish: a challenge for hatchery-based recovery
efforts. Can. J. Fish. Aquat. Sci. 65, 401–408.
https://doi.org/10.1139/f07-176.
35. Krabbenhoft, T.J., MacGuigan, D.J.,
Backenstose, N.J.C., Waterman, H., Lan, T.,
Pelosi, J.A., Tan, M., and Sandve, S.R. (2021).
Chromosome-level genome assembly of
Chinese sucker (Myxocyprinus asiaticus)
reveals strongly conserved synteny following
a Catostomid-specific whole-genome
duplication. Genome Biol. Evol. 13, evab190.
https://doi.org/10.1093/gbe/evab190.
36. Chen, L.-Y., Morales-Briones, D.F., Passow,
C.N., and Yang, Y. (2019). Performance of
gene expression analyses using de novo
assembled transcripts in polyploid species.
Bioinformatics 35, 4314–4320. https://doi.
org/10.1093/bioinformatics/btz620.
37. Lian, S., Zhou, Y., Liu, Z., Gong, A., and
Cheng, L. (2020). The differential expression
patterns of paralogs in response to stresses
indicate expression and sequence
divergences. BMC Plant Biol. 20, 277. https://
doi.org/10.1186/s12870-020-02460-x.
38. Dew, G., Murphy, G., Stanton, H., Vallon, R.,
Angel, P., Reynolds, J.J., and Hembry, R.M.
(2000). Localisation of matrix
metalloproteinases and TIMP-2 in resorbing
mouse bone. Cell Tissue Res. 299, 385–394.
https://doi.org/10.1007/s004419900166.
39. Mosig, R.A., Dowling, O., DiFeo, A., Ramirez,
M.C.M., Parker, I.C., Abe, E., Diouri, J., Aqeel,
A.A., Wylie, J.D., Oblander, S.A., et al. (2007).
Loss of MMP-2 disrupts skeletal and
ll
OPEN ACCESS
iScience 27, 109566, April 19, 2024 11
iScienc
e
Article

craniofacial development and results in
decreased bone mineralization, joint erosion
and defects in osteoblast and osteoclast
growth. Hum. Mol. Genet. 16, 1113–1123.
https://doi.org/10.1093/hmg/ddm060.
40. Hillegass, J.M., Villano, C.M., Cooper, K.R.,
and White, L.A. (2008). Glucocorticoids alter
craniofacial development and increase
expression and activity of matrix
metalloproteinases in developing zebrafish
(Danio rerio). Toxicol. Sci. 102, 413–424.
https://doi.org/10.1093/toxsci/kfn010.
41. Ahi, E.P., Guðbrandsson, J., Kapralova, K.H.,
Franzdo
´ttir, S.R., Snorrason, S.S., Maier, V.H.,
and Jo
´nsson, Z.O. (2013). Validation of
reference genes for expression studies
during craniofacial cevelopment in Arctic
charr. PLoS One 8, e66389. https://doi.org/
10.1371/journal.pone.0066389.
42. Kawauchi, S., Calof, A.L., Santos, R., Lopez-
Burks, M.E., Young, C.M., Hoang, M.P., Chua,
A., Lao, T., Lechner, M.S., Daniel, J.A., et al.
(2009). Multiple organ system defects and
transcriptional dysregulation in the Nipbl+/
mouse, a model of Cornelia de Lange
Syndrome. PLoS Genet. 5, e1000650. https://
doi.org/10.1371/journal.pgen.1000650.
43. Smith, T.G., Laval, S., Chen, F., Rock, M.J.,
Strachan, T., and Peters, H. (2014). Neural
crest cell-specific inactivation of Nipbl or
Mau2 during mouse development results in a
late onset of craniofacial defects. Genesis 52,
687–694. https://doi.org/10.1002/dvg.22780.
44. Gu, W., Wang, L., Gu, R., Ouyang, H., Bao, B.,
Zheng, L., and Xu, B. (2021). Defects of
cohesin loader lead to bone dysplasia
associated with transcriptional disturbance.
J. Cell. Physiol. 236, 8208–8225. https://doi.
org/10.1002/jcp.30491.
45. Altschul, S.F., Gish, W., Miller, W., Myers,
E.W., and Lipman, D.J. (1990). Basic local
alignment search tool. J. Mol. Biol. 215,
403–410. https://doi.org/10.1016/S0022-
2836(05)80360-2.
46. Stankowski, S., and Ravinet, M. (2021).
Defining the speciation continuum. Evolution
75, 1256–1273. https://doi.org/10.1111/evo.
14215.
47. Jorna, J., Linde, J.B., Searle, P.C., Jackson,
A.C., Nielsen, M.-E., Nate, M.S., Saxton, N.A.,
Grewe, F., Herrera-Campos, M.d.l.A., Spjut,
R.W., et al. (2021). Species boundaries in the
messy middle—A genome-scale validation of
species delimitation in a recently diverged
lineage of coastal fog desert lichen fungi.
Ecol. Evol. 11, 18615–18632. https://doi.org/
10.1002/ece3.8467.
48. Brimhall, W.H., and Merritt, L.B. (1981).
Geology of Utah Lake: Implications for
resource management. Great Basin Nat.
Mem. 24–42.
49. Belk, M.C., and Schaalje, G.B. (2016).
Multivariate heritability of shape in June
sucker (Chasmistes liorus) and Utah sucker
(Catostomus ardens): shape as a functional
trait for discriminating closely related species.
Dev. Gene. Evol. 226, 197–207. https://doi.
org/10.1007/s00427-016-0547-2.
50. Cole, D.D. (2014). Ecomorphological and
genetic investigations into the Utah Lake, UT
sucker complex with comparisons to the
Jackson Lake, WY sucker complex. Doctor of
Philosophy - Ecology (Utah State University).
51. Dowling, T.E., Markle, D.F., Tranah, G.J.,
Carson, E.W., Wagman, D.W., and May, B.P.
(2016). Introgressive hybridization and the
evolution of lake-adapted catostomid fishes.
PLoS One 11, e0149884. https://doi.org/10.
1371/journal.pone.0149884.
52. Suetsugu, K., Fukushima, K., Makino, T.,
Ikematsu, S., Sakamoto, T., and Kimura, S.
(2023). Transcriptomic heterochrony and
completely cleistogamous flower
development in the mycoheterotrophic
orchid Gastrodia. New Phytol. 237, 323–338.
https://doi.org/10.1111/nph.18495.
53. Ha
¨rer, A., Torres-Dowdall, J., and Meyer, A.
(2017). Rapid adaptation to a novel light
environment: The importance of ontogeny
and phenotypic plasticity in shaping the
visual system of Nicaraguan Midas cichlid fish
(Amphilophus citrinellus spp.). Mol. Ecol. 26,
5582–5593. https://doi.org/10.1111/mec.
14289.
54. Vea, I.M., and Minakuchi, C. (2021). Atypical
insects: molecular mechanisms of unusual life
history strategies. Curr. Opin. Insect Sci. 43,
46–53. https://doi.org/10.1016/j.cois.2020.
09.016.
55. Johnson, C.K., and Voss, S.R. (2013). Chapter
Eight - Salamander Paedomorphosis: Linking
Thyroid Hormone to Life History and Life
Cycle Evolution. In Current Topics in
Developmental Biology, Y.-B. Shi, ed.
(Academic Press), pp. 229–258. https://doi.
org/10.1016/B978-0-12-385979-2.00008-3.
56. Gillis, J.A., Dahn, R.D., and Shubin, N.H.
(2009). Chondrogenesis and homology of the
visceral skeleton in the little skate, Leucoraja
erinacea (Chondrichthyes: Batoidea).
J. Morphol. 270, 628–643. https://doi.org/10.
1002/jmor.10710.
57. Thisse, C., and Thisse, B. (2008). High-
resolution in situ hybridization to whole-
mount zebrafish embryos. Nat. Protoc.
3, 59–69.
58. Searle, P.C. (2024). Pcsearle/Sucker: Sucker
v1.1.0 (v1.1.0). https://doi.org/10.5281/
zenodo.10806437.
59. Rohlf, F.J. (2015). The tps series of software.
Hystrix 26, 9–12. https://doi.org/10.4404/
hystrix-26.1-11264.
60. Salmela, L., and Rivals, E. (2014). LoRDEC:
Accurate and efficient long read error
correction. Bioinformatics 30, 3506–3514.
https://doi.org/10.1093/bioinformatics/
btu538.
61. Li, W., Fu, L., Niu, B., Wu, S., and Wooley, J.
(2012). Ultrafast clustering algorithms for
metagenomic sequence analysis. Briefings
Bioinf. 13, 656–668. https://doi.org/10.1093/
bib/bbs035.
62. Buchfink, B., Reuter, K., and Drost, H.-G.
(2021). Sensitive protein alignments at tree-
of-life scale using DIAMOND. Nat. Methods
18, 366–368. https://doi.org/10.1038/s41592-
021-01101-x.
63. Go
¨tz, S., Garcı
´a-Go
´mez, J.M., Terol, J.,
Williams, T.D., Nagaraj, S.H., Nueda, M.J.,
Robles, M., Talo
´n, M., Dopazo, J., and
Conesa, A. (2008). High-throughput
functional annotation and data mining with
the Blast2GO suite. Nucleic Acids Res. 36,
3420–3435. https://doi.org/10.1093/nar/
gkn176.
64. Jones, P., Binns, D., Chang, H.-Y., Fraser, M.,
Li, W., McAnulla, C., McWilliam, H., Maslen,
J., Mitchell, A., Nuka, G., et al. (2014).
InterProScan 5: genome-scale protein
function classification. Bioinformatics 30,
1236–1240. https://doi.org/10.1093/
bioinformatics/btu031.
65. Cantalapiedra, C.P., Herna
´ndez-Plaza, A.,
Letunic, I., Bork, P., and Huerta-Cepas, J.
(2021). eggNOG-mapper v2: functional
annotation, orthology assignments, and
domain prediction at the metagenomic scale.
Mol. Biol. Evol. 38, 5825–5829. https://doi.
org/10.1093/molbev/msab293.
66. Huerta-Cepas, J., Forslund, K., Coelho, L.P.,
Szklarczyk, D., Jensen, L.J., von Mering, C.,
and Bork, P. (2017). Fast genome-wide
functional annotation through orthology
assignment by eggNOG-mapper. Mol. Biol.
Evol. 34, 2115–2122. https://doi.org/10.1093/
molbev/msx148.
67. Huerta-Cepas, J., Szklarczyk, D., Heller, D.,
Herna
´ndez-Plaza, A., Forslund, S.K., Cook, H.,
Mende, D.R., Letunic, I., Rattei, T., Jensen,
L.J., et al. (2019). eggNOG 5.0: a hierarchical,
functionally and phylogenetically annotated
orthology resource based on 5090 organisms
and 2502 viruses. Nucleic Acids Res. 47,
D309–D314. https://doi.org/10.1093/nar/
gky1085.
68. Krueger, F. (2012). Trim Galore. https://www.
bioinformatics.babraham.ac.uk/projects/
trim_galore/.
69. Andrews, S. (2010). FastQC: A quality control
tool for high throughput sequence data.
http://www.bioinformatics.babraham.ac.uk/
projects/fastqc/.
70. Patro, R., Duggal, G., Love, M.I., Irizarry, R.A.,
and Kingsford, C. (2017). Salmon provides
fast and bias-aware quantification of
transcript expression. Nat. Methods 14,
417–419. https://doi.org/10.1038/
nmeth.4197.
71. Soneson, C., Love, M.I., and Robinson, M.D.
(2015). Differential analyses for RNA-seq:
Transcript-level estimates improve gene-
level inferences [version 2; peer review: 2
approved]. F1000Res. 4, 1521. https://doi.
org/10.12688/f1000research.7563.2.
72. Love, M.I., Huber, W., and Anders, S. (2014).
Moderated estimation of fold change and
dispersion for RNA-seq data with DESeq2.
Genome Biol. 15, 550. https://doi.org/10.
1186/s13059-014-0550-8.
73. Belk, M.C., Maxwell, M., Laidlaw, C., and
Wesner, J. (2016). Building a better June
sucker: characterization of mouth shape in
the captive brood stock. Open Fish Sci. J.
9, 29–36.
74. Rohlf, F.J. (2003). tpsDig: Digitize
Landmarks & Outlines from Image Files,
Scanner, or Video (Department of Ecology
and Evolution, State University of New York at
Stony Brook).
75. Bookstein, F.L. (1997). Landmark methods for
forms without landmarks: Morphometrics of
group differences in outline shape. Med.
Image Anal. 1, 225–243. https://doi.org/10.
1016/S1361-8415(97)85012-8.
76. Rohlf, F.J., Loy, A., and Corti, M. (1996).
Morphometric analysis of Old World
Talpidae (Mammalia, Insectivora) using
partial-warp scores. Syst. Biol. 45, 344–362.
https://doi.org/10.1093/sysbio/45.3.344.
77. Love, M.I., Hogenesch, J.B., and Irizarry, R.A.
(2016). Modeling of RNA-seq fragment
sequence bias reduces systematic errors in
transcript abundance estimation. Nat.
Biotechnol. 34, 1287–1291. https://doi.org/
10.1038/nbt.3682.
78. Carbon, S., Ireland, A., Mungall, C.J., Shu, S.,
Marshall, B., and Lewis, S.; AmiGO Hub; Web
Presence Working Group (2009). AmiGO:
Online access to ontology and annotation
data. Bioinformatics 25, 288–289. https://doi.
org/10.1093/bioinformatics/btn615.
79. Ashburner, M., Ball, C.A., Blake, J.A.,
Botstein, D., Butler, H., Cherry, J.M., Davis,
A.P., Dolinski, K., Dwight, S.S., Eppig, J.T.,
et al. (2000). Gene ontology: tool for the
ll
OPEN ACCESS
12 iScience 27, 109566, April 19, 2024
iScienc
e
Article

unification of biology. Nat. Genet. 25, 25–29.
https://doi.org/10.1038/75556.
80. Gene Ontology Consortium (2021). The Gene
ontology resource: enriching a GOld mine.
Nucleic Acids Res. 49, D325–D334. https://
doi.org/10.1093/nar/gkaa1113.
81. Rohlf, F.J. (2003). tpsRelw32: Relative Warps
Analysis (Department of Ecology and
Evolution, State University of New York at
Stony Brook).
82. Williams, T.J., Johnson, J.B., and Belk, M.C.
(2017). Interaction between predation
environment and diet constrains body shape
in Utah chub, Gila atraria (Cypriniformes:
Cyprinidae). Biol. J. Linn. Soc. Lond. 122,
147–156. https://doi.org/10.1093/biolinnean/
blx050.
83. Roth-Monzo
´n, A.J., Belk, M.C., Zu
´n
˜iga-Vega,
J.J., and Johnson, J.B. (2020). Beyond
pairwise interactions: Multispecies character
displacement in Mexican freshwater fish
communities. Am. Nat. 195, 983–996. https://
doi.org/10.1086/708513.
84. Searle, P.C., Mercer, M., Habit, E., and Belk,
M.C. (2021). Ontogenetic shape trajectory of
Trichomycterus areolatus varies in response
to water velocity environment. PLoS One 16,
e0252780. https://doi.org/10.1371/journal.
pone.0252780.
85. Mercer, M., Searle, P.C., Cifuentes, R., Habit,
E., and Belk, M.C. (2020). Morphometric
response of Galaxias maculatus (Jenyns) to
lake colonization in Chile. Diversity 12, 219.
https://doi.org/10.3390/d12060219.
86. Wesner, J.S., Billman, E.J., Meier, A., and
Belk, M.C. (2011). Morphological
convergence during pregnancy among
predator and nonpredator populations of the
livebearing fish Brachyrhaphis rhabdophora
(Teleostei: Poeciliidae). Biol. J. Linn. Soc.
Lond. 104, 386–392. https://doi.org/10.1111/
j.1095-8312.2011.01715.x.
87. Hassell, E.M.A., Meyers, P.J., Billman, E.J.,
Rasmussen, J.E., and Belk, M.C. (2012).
Ontogeny and sex alter the effect of
predation on body shape in a livebearing fish:
Sexual dimorphism, parallelism, and costs of
reproduction. Ecol. Evol. 2, 1738–1746.
https://doi.org/10.1002/ece3.278.
88. Ingley, S.J., Billman, E.J., Belk, M.C., and
Johnson, J.B. (2014). Morphological
divergence driven by predation environment
within and between species of Brachyrhaphis
fishes. PLoS One 9, e90274. https://doi.org/
10.1371/journal.pone.0090274.
89. Kenward, M.G., and Roger, J.H. (1997). Small
sample inference for fixed effects from
restricted maximum likelihood. Biometrics
53, 983–997. https://doi.org/10.2307/
2533558.
90. Baken, E.K., Collyer, M.L., Kaliontzopoulou,
A., and Adams, D.C. (2021). geomorph v4.0
and gmShiny: Enhanced analytics and a new
graphical interface for a comprehensive
morphometric experience. Methods Ecol.
Evol. 12, 2355–2363. https://doi.org/10.1111/
2041-210X.13723.
91. Adams, D.C., Collyer, M.L., Kaliontzopoulou,
A., and Baken, E.K. (2021). Geomorph:
Software for geometric morphometric
analyses. R package version 4.0.3. https://
cran.r-project.org/package=geomorph.
92. R Development Core Team (2020). R: A
Language and Environment for Statistical
Computing (R Foundation for Statistical
Computing).
93. Langerhans, R.B. (2009). Trade-off between
steady and unsteady swimming underlies
predator-driven divergence in Gambusia
affinis. J. Evol. Biol. 22, 1057–1075. https://
doi.org/10.1111/j.1420-9101.2009.01716.x.
94. Langerhans, R.B., and Makowicz, A.M. (2009).
Shared and unique features of morphological
differentiation between predator regimes in
Gambusia caymanensis. J. Evol. Biol. 22,
2231–2242. https://doi.org/10.1111/j.1420-
9101.2009.01839.x.
ll
OPEN ACCESS
iScience 27, 109566, April 19, 2024 13
iScienc
e
Article

STAR+METHODS
KEY RESOURCES TABLE
REAGENT or RESOURCE SOURCE IDENTIFIER
Biological samples
Chasmistes liorus State of Utah Fisheries Experiment Station See Table S1.
Catostomus ardens (2 lineages) State of Utah Fisheries Experiment Station See Table S1.
Chemicals, peptides, and recombinant proteins
Tricaine Methanesulfonate MilliporeSigma Catalog #: E10521
RNALater MilliporeSigma Catalog #: R0901
Modified Carnoy’s solution N/A N/A
Critical commercial assays
Alizarin Red/Alcian Blue Staining Protocol Gillis et al.,
56
Thisse and Thisse
57
N/A
TRIzol Plus RNA Purification Kit Thermo Fisher Scientific Catalog #: 12183555
ProNex Size-Selective Purification System Promega Catalog #: NG2001
Iso-Seq Express 2.0 Workflow Pacific Biosciences N/A
RNeasy Plus Mini Kit Qiagen Catalog #: 74314
NEBNext Ultra II RNA with Poly-A selection New England Biolabs Catalog #: E7775
Deposited data
Genetic Data This Paper NCBI: PRJNA818778
Morphometric Data This Paper
58
Zenodo: https://zenodo.org/records/10806437
Code This Paper
58
Zenodo: https://zenodo.org/records/10806437
Software and algorithms
tpsUtil, tpsDig, tpsRelw Rohlf
59
https://life2.bio.sunysb.edu/ee/rohlf/software.html
SAS SAS Institute Inc. v9.4; https://www.sas.com/en_us/home.html
R R Development Core Team v4.1.1; https://www.r-project.org/
Python N/A v. 3.11.5
SMRT Link Run QC Pacific Biosciences v9.0; https://www.pacb.com/support/
software-downloads/
IsoSeq3 Pacific Biosciences v3.4.0; https://github.com/PacificBiosciences/IsoSeq
LoRDEC Salmela and Rivals
60
v0.9
CD-HIT-Est Li et al.
61
v4.8.1
DIAMOND Buchfink et al.
62
v2.0.11
BUSCO Seppey et al.
25
v5.1.2
OmicsBox BioBam v2.0.24; https://www.biobam.com/
Blast2GO Go
¨tz et al.
63
N/A
InterProScan Jones et al.
64
v86.0
eggNOG-mapper, eggNOG Cantalapiedra et al.
65
Huerta-Cepas et al.,
66
Huerta-Cepas et al.,
67
v2.1.0, v5.0.2
TrimGalore Krueger
68
v0.6.5
FastQC Andrews
69
v0.11.9
Salmon Patro et al.
70
v1.8.0
tximport Soneson et al.
71
v1.26.0
DESeq2 Love et al.
72
v1.32.0
(Continued on next page)
ll
OPEN ACCESS
14 iScience 27, 109566, April 19, 2024
iScienc
e
Article

RESOURCE AVAILABILITY
Lead contact
Further information and requests for resources should be directed to and will be fulfilled by the lead contact: Peter C. Searle (pcs222@
cornell.edu).
Materials availability
The study did not generate new unique reagents.
Data and code availability
All morphometric data have been deposited at Zenodo
58
and are publicly available as of the date of publication. DOIs are listed in the
key resources table. All genetic data associated with the publication are found at NCBI: PRJNA818778 and are publicly available as of
the date of publication.
All original code has been deposited at Zenodo
58
and is publicly available as of the date of publication. DOIs are listed in the key re-
sources table.
Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS
June sucker and Utah sucker
June sucker—once abundant in Utah Lake—declined to a historical low of approximately 1,000 individuals in 1989. During the 1990’s and early
2000’s, captive brood lots were created from forced spawning of wild June sucker to augment natural reproduction while critical spawning
habitat was restored.
73
Belk et al.
73
formally evaluated lower lip morphology of these brood lots and determined that lower lip morphology
ranged between that of pure June sucker (crosses between wild June sucker from Utah Lake; thin, separated lower lip lobes with few papillae)
and pure Utah sucker (crosses between wild Utah sucker from an introduced population in Strawberry Reservoir, Utah, USA; broad lower lip
lobes with no gap between them, that are highly papillose), with most brood lots exhibiting intermediate phenotypes.
49
For our analysis, we
acquired samples from two of these brood lots from Utah Lake: a lot identified as the most June-sucker-like (June sucker) and another lot
identified as the most Utah-sucker-like (Utah Lake Utah sucker). We also obtained larval samples originating from wild Utah suckers from
Strawberry Reservoir (Strawberry Reservoir Utah sucker). We included fish from Strawberry Reservoir so we could analyze development in
Utah suckers that were not influenced by suckers from Utah Lake.
The three sucker lineages were hatched and reared in 2014 at the State of Utah Fisheries Experiment Station (Logan, UT, USA) in a
common-garden environment. After hatching, larval samples were collected for 15-weeks: once weekly for weeks 0-2, twice weekly for
weeks 3-11, and again once weekly for weeks 12-14 (n=416) following approved Brigham Young University Institutional Animal Care
and Use Committee (IACUC) protocols (IACUC -approved protocol #15-0602, Utah Division of Wildlife Resources’ scientific collection
permit #1COLL5950). Specimens collected on week 0 were 0 days after hatching, whereas specimens collected on week 14 were
98 days after hatching. At each sampling date, twelve fish were collected from each sucker lineage and euthanized with tricaine metha-
nesulfonate (MS-222, MilliporeSigma, St. Louis, MO, USA). Six specimens were stored in RNAlater (MilliporeSigma, St. Louis, MO, USA) at
-20C for RNA sequencing. Six specimens were stored in a modified Carnoy’s solution (60 mL 100% ethanol, 30 mL 37% formaldehyde,
10 mL glacial acetic acid) at 4C for 1 to 4 days for the geometric morphometric analysis. The fish stored in the modified Carnoy’s solution
were then dehydrated in 75% ethanol (75 mL 100% ethanol, 25 mL diH
2
O; three 30 min intervals), 100% ethanol (3 one-hour intervals) and
stored at -20C. Because these specimens were in the larval stage of development, sex could not be determined and was not included in
analyses.
METHOD DETAILS
Geometric morphometrics
We photographed the left lateral head of all specimens (n=416; 140 June sucker, 143 Utah Lake Utah sucker, 133 Strawberry Reservoir Utah
sucker) using an Olympus stereo microscope (Olympus, Shinjuku City, Tokyo, Japan). Eighteen fish were misshapen and were removed from
our analysis (4 June sucker, 9 Utah Lake Utah sucker, and 5 Strawberry Reservoir Utah sucker). Strawberry Reservoir Utah sucker collected
Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
Other
Olympus stereo microscope Olympus N/A
PacBio Sequel II System Pacific Biosciences N/A
Illumina NovaSeq 6000 System Illumina N/A
ll
OPEN ACCESS
iScience 27, 109566, April 19, 2024 15
iScienc
e
Article

during week 2 were misplaced and thus, were not included in the analysis. After these steps, 398 fish remained including 136 June sucker,
134 Utah Lake Utah sucker, and 128 Strawberry Reservoir Utah sucker (Table S1).
To photograph the premaxilla, a subset of the formalin fixed fish (n=213) was cleared and stained using a modified Alizarin Red/Alcian
Blue staining protocol.
56
Prior to clearing and staining, pigment was removed using a hydrogen peroxide treatment (.5g KOH, 90 mL 1x
PBS, 10 mL 30% H
2
O
2)
. Fish were incubated in the hydrogen peroxide solution at room temperature until pigment disappeared and then
washed in series (1X PBS, 25%, 50% and 75% ethanol in 1X PBS, 10 min/wash).
57
Next, the fish were placed in 70% ethanol (30 mL diH
2
O,
70 mL 100% ethanol, 2 h) and then acetic ethanol (30 mL glacial acetic acid, 70 mL 100% ethanol, 2 h). Cartilage was stained using Alcian
Blue 8GX (MilliporeSigma, St. Louis, MO, USA; 20 mg Alcian Blue, 100 mL acetic ethanol, 24 h). Specimens were destained in acetic
ethanol and rehydrated in series (70%, 50%, 25% ethanol in diH
2
O and 100% diH
2
O, 2 h). Tissue was cleared using a trypsin solution
(Ward’s Science, St. Catharines, ON, Canada; 1 g trypsin, 2 g sodium tetraborate, 97 mL diH
2
O) until approximately 50% clear and
then placed in an Alizarin Red S solution (MilliporeSigma, St. Louis, MO, USA; 0.1 g Alizarin Red S, 100 mL 1% KOH, 24 h) to stain
bone. Specimens were again cleared in a trypsin solution until 100% clear and placed in a .5% KOH solution (75 mL .5% KOH, 25 mL glyc-
erol, 24 h). Specimens were soaked in a .5% KOH/glycerol series (3:1, 1:1, 1:3, .5% KOH:glycerol, 24 h/wash) and stored in 80% glycerol
(80 mL glycerol, 20 mL 1x PBS).
After staining, we photographed the premaxilla of 213 fish (72 June sucker, 72 Utah Lake Utah sucker and 69 Strawberry Reservoir Utah
sucker) using an Olympus stereo microscope. The premaxilla for weeks 0-2 was not developed enough to photograph and was excluded
from our analysis (n=24). The premaxilla of one fish was damaged and was excluded from our analysis (n=1). After these steps, 188 fish re-
mained including 63 June sucker, 63 Utah Lake Utah sucker, and 62 Strawberry Reservoir Utah sucker (Table S1).
We used tpsDig to digitize landmarks.
74
To ensure consistent, homologous placement of landmarks, we had one researcher place land-
marks and another researcher independently inspect the landmarks. For the head, we included 12 landmarks: (1) posterior maximum of oper-
culum, (2) posterior maximum of eye, (3) anterior maximum of eye, (4) dorsal margin of upper lip, (5) anterior maximumof eye projected onto
the dorsal outline, (6) posterior maximum of eye projected onto the dorsal outline, (7) posterior maximum of operculum projected onto the
dorsal outline, (8) posterior maximum of operculum projected onto the ventral outline, (9) posterior maximum of eye projected onto the
ventral outline, (10) anterior maximum of eye projected onto the ventral outline, (11) midpoint between landmarks 4 and 10 on the ventral
outline, and (12) midpoint between landmarks 4 and 5 on the dorsal outline. Landmarks 11 and 12 were sliding semi-landmarks (Figure S3;
sensu
75
). For the premaxilla, we included 8 landmarks: (1) maximum of vertical limb, (2) maximum of medial angle, (3) maximum of lateral limb,
(4) midpoint between 1 and 3 on right lateral outline (5) midpoint between 1 and 4 on right lateral outline, (6) midpoint between 1 and 2 on
medial outline, (7) midpoint between 2 and 3 on ventral outline, and (8) midpoint between 3 and 4 on right lateral outline. Landmarks 4
through 8 were sliding semi-landmarks (Figure S4). The premaxilla is a symmetrical structure, so we only landmarked the left premaxilla of
each specimen.
76
See description of statistical analyses below.
Sequencing
We generated PacBio long-reads for de novo transcriptome assembly. For PacBio sequencing, specimens were thawed and placed in
grooves on agar plates made with custom 3D molds, ensuring consistent placement of specimens for tissue removal. The entire head of
each specimen was excised by cutting directly behind the operculum with a scalpel. A new groove and sanitized scalpel were used for
each specimen. Three pooled samples consisting of 3 specimens per lineage—1 specimen from weeks 8, 11, and 13—were used for
PacBio sequencing (Table S1). Weeks 9, 12, and 14 were identified as occurring during the critical developmental period from our geometric
morphometric analyses (see Results) but specimens were used from weeks 8, 11, and 13 instead of weeks 9, 12, and 14 because not enough
specimens were available from weeks 9, 12, and 14 for both PacBio and Illumina sequencing. We used the whole head of specimens from
multiple weeks to generate a comprehensive transcriptome represented by multiple tissues. Tissues were frozen in RNAlater
(MilliporeSigma, St. Louis, MO, USA) and delivered to Brigham Young Universities’ DNA Sequencing Center (Provo, UT, USA) on dry ice
for RNA extraction, library preparation and sequencing. Total RNA was extracted according to the manufacturer’s instructions using the TRI-
zol Plus RNA Purification kit (Thermo Fisher Scientific, Waltham, MA, USA), secondary cleanup was performed using the ProNex Size-Selective
Purification System (Promega, Madison, WI, USA) and total RNA was then pooled into a single sample for each lineage (as described above).
Barcoded cDNA libraries were prepared according to the Iso-Seq Express 2.0 workflow system (Pacific Biosciences, Menlo Park, CA, USA)
using polyA selection. Libraries were sequenced using a PacBio Sequel II System on a single SMRT cell 8M for a 24-h movie with insert
size of 2,500 bp (Pacific Biosciences, Menlo Park, CA, USA).
We generated Illumina short-reads for error correction of our transcriptomes, as well as RNA-seq analyses. For Illumina sequencing, tissue
at the tip of the rostrum, directly in front of the eye, was removed (as described above). Tissues were frozen in RNAlater and shipped to Ad-
mera Health, LLC (South Plainfield, NJ, USA) on dry ice for RNA extraction, library preparation, and sequencing. Total RNA was successfully
extracted from 34 specimens from weeks 9 (n=12), 12 (n=12), and 14 (n=10) for Illumina sequencing. For week 14, total RNA was only success-
fully extracted from 10 specimens (June sucker, n=3; Utah Lake Utah sucker, n=4; Strawberry Reservoir Utah sucker, n=3, Table S1). Total RNA
was extracted according to the manufacturer’s instructions using a RNeasy Plus Mini kit (Qiagen, Hilden, Germany). Barcoded cDNA libraries
were prepared from total RNA using the NEBNext Ultra II RNA with Poly-A selection kit (New England Biolabs, Ipswich, MA, USA) which se-
lects for mRNA. Libraries were sequenced using an Illumina NovaSeq 6000 system with S4 flow cell and read length of 2x150 bp (Illumina, San
Diego, CA, USA).
ll
OPEN ACCESS
16 iScience 27, 109566, April 19, 2024
iScienc
e
Article

Transcriptome assembly
PacBio sequence quality was checked using SMRT Link Run QC (v9.0; (https://www.pacb.com/support/software-downloads/) before the se-
quences were processed using the IsoSeq3 (v3.4.0) protocol (https://github.com/PacificBiosciences/IsoSeq, See Figure S5 for diagram out-
lining key steps for assembly, annotation, and differential expression analyses). In brief, circular consensus sequences were generated using
ccs (v6.0.0); primers were removed, and barcodes identified using lima (v2.0.0); poly (A) tails trimmed and concatemers identified and
removed using IsoSeq3 refine (v3.4.0); and full-length non-concatemer reads clustered using IsoSeq3 cluster (v3.4.0). High- and low-quality
transcripts were identified using a predicted accuracy cutoff of 0.99 and only high-quality transcripts were retained for subsequent analyses
(Table S2A). High quality transcripts were error corrected using all Illumina sequences from each lineage (pre-processing of Illumina se-
quences described below) using LoRDEC (v0.9).
60
Finally, error-corrected transcripts were clustered using CD-HIT-Est (v4.8.1) to remove addi-
tional sequence redundancy (Table S2B).
61
Transcripts were checked for contamination by blasting the transcripts against the nr database (NCBI non-redundant protein sequences,
https://www.ncbi.nlm.nih.gov/) using DIAMOND (v2.0.11).
62
Top 10 hits using DIAMOND all came from fish in the order Cypriniformes (June
sucker, 91.78%; Utah Lake Utah sucker, 90.89%; Strawberry Reservoir Utah sucker, 91.91%) indicating no significant contamination, so all tran-
scripts were retained for future analyses. BUSCO (Benchmarking Universal Single-Copy Orthologs; v5.1.2; -l actinopterygii_odb10) was used
to assess transcriptome assembly completeness.
25
Functional annotation
We functionally annotated our transcriptomes to validate that the transcriptomes were similar between lineages. We used DIAMOND (as
described above) to obtain the best blast hits for each transcript from the nr database (NCBI non-redundant protein sequences, https://
www.ncbi.nlm.nih.gov/). Blast hits and transcripts were then imported into OmicsBox (v2.0.24, https://www.biobam.com/). Next, candidate
gene ontology terms were retrieved for each transcript and assigned to the transcripts using Blast2GO.
63
We obtained additional gene
ontology terms using InterProScan5 (v86.0), which searches the InterPro member databases using protein signature and protein domain in-
formation,
64
as well as eggNOG-mapper (v2.1.0) which uses precalculated orthologous groups from the eggNOG (v5.0.2) database.
65–67
These additional gene ontology terms were merged with the Blast2GO annotations.
We generated common transcriptomes that consisted only of shared transcripts among all lineages so we could compare differential
expression patterns between lineages. We reran eggNOG-mapper (v2.1.12), this time restricting annotations to species within Actinopterygii
to generate as fine-scale annotations as possible. We used the predicted gene names generated from this program to then filter the tran-
scripts using R. EggNOG-mapper identified predicted gene names for 21,033 (57.1%), 19,679 (55.2%), and 18,255 (56.8%) transcripts for
June sucker, Utah Lake Utah sucker, and Strawberry Reservoir Utah sucker, respectively. After filtering, 18,726, 17,849, and 16,787 tran-
scripts—consisting of 8,392 common genes among all lineages—remained for June sucker, Utah Lake Utah sucker, and Strawberry Reservoir
Utah sucker, respectively. Initially, when multiple transcripts shared the same gene name, we only retained the transcript with the lowest
e-value for each lineage. However, we found that this final filtering step often led to the selection of transcripts that were dissimilar
(< 97% identity) between lineages, despite lineages sharing highly similar transcripts (>99.5% identity) prior to filtering. To avoid using dis-
similar transcripts between lineages we performed additional clustering. Specifically, we a) merged all the transcripts identified as each gene
for all the sucker lineages, b) clustered these transcripts using CD-HIT-Est, and c) only retained clusters of transcripts that were represented by
all lineages and had a cluster identity greater than or equal to 97%. Also, when multiple transcripts were present for a lineage in a cluster, we
only retained the transcript with the highest cluster identity. This additional clustering step resulted in transcriptomes consisting of 7,195
genes and 10,426 transcripts for each lineage. In these transcriptomes, the number of transcripts associated with the genes ranged from
1 to 11 transcripts, with most genes only having one transcript.
Differential expression
We generated count tables using our assembled and annotated transcriptomes, as well as the Illumina sequences. Specifically, we pre-pro-
cessed the Illumina sequences by performing read adapter and quality trimming with Trim Galore (v0.6.5) and then checked sequence quality
using FastQC (v.0.11.9).
68,69
Transcript-level abundances were generated with Salmon (v1.8.0).
70,77
The overall alignment rate was 35.40%,
33.06%, and 33.41% for June sucker, Utah Lake Utah sucker, and Strawberry Reservoir Utah sucker, respectively. We used tximport
(v1.26.0) to generate transcript-level count matrices.
71
We chose to analyze transcripts annotated as the same gene at the transcript-level
so we could explore how expression changed between these divergent transcripts. All subsequent transcript expression analyses were per-
formed using these count matrices with DESeq2 (v1.32.0) in R.
72
The quality of the count tables were assessed prior to running differential expression analyses. Transcript-level counts were normalized and
transformed using the regularized logarithm (rlog) in DESeq2 and analyzed using a separate principal components analysis for each lineage.
The principal components analysis for each sucker lineage revealed two potential outliers for June sucker and Strawberry Reservoir Utah
sucker, both in week 12 (Figure S6). Although the fish were raised in common garden conditions, differences in preservation and relative
age (i.e., not all fish were the same length when sampled) could have contributed to these differences. Analyses were conducted including
and excluding these samples. General patterns (as described in the Results) were consistent with and without these samples, so these two
samples were included in all subsequent analyses. See description of statistical analyses below.
ll
OPEN ACCESS
iScience 27, 109566, April 19, 2024 17
iScienc
e
Article

Candidate genes
An initial list of candidate genes was generated with AmiGO2
78
on Geneontology.org
79,80
using two separate searches (‘‘development+face’’
and ‘‘development+craniofacial’’). From these searches, 1059 gene annotations (332 unique genes) were identified, of which 79 transcripts
were present in our filtered transcriptomes. Of these 79 candidate transcripts, four transcripts with low counts were removed because DESeq2
was unable to determine if the transcripts were differentially expressed. In addition, 25 transcripts were removed that were classified as being
unchanged across all contrasts and lineages, resulting in 50 candidate transcripts (Table 2). Finally, we generated plots of transcript-level
expression (normalized counts) for each candidate transcript using DESeq2 (Figures S7 and S8). The genetic data associated with these an-
alyses are found at NCBI: PRJNA818778 and the original code associated with these analyses has been deposited at Zenodo.
58
QUANTIFICATION AND STATISTICAL ANALYSIS
Geometric morphometrics
We used tpsRelw to generate shape variables from the landmark data.
81
We used the minimize d
2
option in tpsRelw for sliding semi-land-
marks.
59
We used the first 9 of 20 relative warps for the head and the first 6 of 12 relative warps for the premaxilla in our analysis because
they captured 98.98% and 96.77% of the shape variation in the head and premaxilla, respectively. We excluded relative warps to account
for loss of dimensionality from the use of sliding semi-landmarks.
82
We also excluded relative warps that individually accounted for < 1%
of shape variation to avoid inflating the real degrees of freedom associated with analysis of shape variation.
82–84
We ran a multivariate linear mixed model to analyze shape change in the head and premaxilla during development in the sucker line-
ages.
82–88
We analyzed head and premaxilla shape separately, using relative warps as the response variable. The multivariate linear mixed
model requires a univariate response variable, so we vectorized the matrix of relative warps so each individual specimen was represented
by 9 rows (head) and 6 rows (premaxilla) of data. Thus, the response variable is transposed from a row in a matrix to a vector column of re-
sponses. We included an index variable as a discrete predictor variable to maintain the original ordering of the relative warps. The index var-
iable preserved the order of the relative warps so that comparisons were made across matching relative warps between lineages.
82–85
Explan-
atory variables in our model included: lineage (discrete predictor; three values), week (discrete predictor; 15 values for head shape; 6 values for
premaxilla shape, described below), the index variable, the interactions between the main effects and the index variable and the three-way
interaction between lineage, week, and the index variable. We did not include centroid size (multivariate measure of shape derived from anal-
ysis in tpsRelW) as a covariate in our analyses because week acted as an index for stage of development and size. Because relative warps result
from a principal components analysis, they have a mean of zero making interpretation of the main effects and index variable by themselves
uninformative. Thus, we used the two-way interactions (lineage*index and week*index) and the three-way interaction (lineage*week*index) to
test whether shape varied between lineages across ontogeny on at least some set of relative warps (Table 1).
82–85
In addition, we included
individual ID as a random effect in our model because each specimen had multiple relative warps as response variables. Degrees of freedom
were estimated using the Kenward-Roger method.
89
For the premaxilla, we combined adjacent weeks (in pairs) to increase our sample size, so
our statistical analysis converged. We ran all analyses with Proc MIXED in SAS (SAS version 9.4, SAS Institute Inc., Cary NC, USA).
Relative warps 1 and 2 accounted for 64.59% and 9.57% of the shape variation in the head, respectively. Relative warps 1 and 2 accounted
for 54.79% and 22.57% of the shape variation in the premaxilla, respectively. We used geomorph (v4.0.4)
90,91
in R (v4.1.1)
92
to generate thin-
plate spline deformations of head and premaxilla shape for the positive and negative extremes of relative warps 1 and 2 to visualize the range
of shape variation represented on each of the first two relative warps (Figures 1A and 2A).
81
To visualize change in head and premaxilla shape
over ontogeny, we plotted least squares means for each lineage across all weeks on relative warps 1 and 2 (Figures 1B and 2B). We plotted
least squares means and 95% confidence intervals for each relative warp for each lineage at week 14, to determine which relative warps
differed between lineages (Figures 1C and 2C). We calculated divergence vectors to visualize shape variation between lineages at week
14.
84,93,94
Specifically, we ran a principal components analysis on the least-squares means for each relative warp for each pairwise comparison
of sucker lineages. We generated divergence scores by summing the products of the first eigenvector and its associated relative warp score
for each specimen. We regressed divergence score against shape (i.e., generalized Procrustes aligned specimens) using geomorph to
generate thin-plate spline representations of the extremes of shape variation between lineages (Figure 1D).
To compare patterns of size change across the developmental sequence in the premaxilla, we measured absolute distances between land-
marks 1 and 2 (vertical limb length), as well as 2 and 3 (lateral limb length) using geomorph. We summed both limb lengths to create an overall
measure of size. We plotted the means and 95% confidence intervals of size across the developmental sequence (Figure 3) for comparison
among lineages. The code and morphometric data associated with these analyses have been deposited at Zenodo.
58
Differential expression
We identified differentially expressed transcripts (DETs) by running the standard DESeq2 pipeline for each lineage using pairwise compar-
isons (contrasts) of weeks 9 and 12, 9 and 14, and 12 and 14 with a Wald test. Transcripts were considered to be differentially expressed if
they had a Benjamini-Hochberg adjusted p-value < .05. Transcripts were considered to be upregulated if they had a log2 fold change greater
than 0 and downregulated if they had a log2 fold change less than 0.
72
Venn diagrams were created to compare DETs between lineages for
each pairwise comparison.
ll
OPEN ACCESS
18 iScience 27, 109566, April 19, 2024
iScienc
e
Article