Purification of Human β- and γ-actin from
Budding Yeast
Brian K. Haarer (0000-0002-4018-222X)
1
, Morgan L. Pimm (0000-0001-6370-1435)
1
, Ebbing P. de Jong (0000-0002-0362-1772)
2
,
David C. Amberg (0000-0002-4709-8848)
1
, and Jessica L. Henty-Ridilla (0000-0002-7203-8791)
1,3
1
Department of Biochemistry & Molecular Biology, SUNY Upstate Medical University, Syracuse, NY, 13210
2
Mass Spectrometry Core Facility, SUNY Upstate Medical University, Syracuse, NY, 13210
3
Department of Neuroscience & Physiology, SUNY Upstate Medical University, Syracuse, NY, 13210
Biochemical studies of human actin and its binding partners
rely heavily on abundant and easily purified α-actin from skele-
tal muscle. Therefore, muscle actin has been used to evaluate
and determine the activities of most actin regulatory proteins
and there is an underlying concern that these proteins perform
differently with actin present in non-muscle cells. To provide
easily accessible and relatively abundant sources of human β- or
γ-actin (i.e., cytoplasmic actins), we developed Saccharomyces
cerevisiae strains that express each as their sole source of actin.
Both β- or γ-actin purified in this system polymerize and in-
teract with various binding partners, including profilin, mDia1
(formin), fascin, and thymosin-β4 (Tβ4). Notably, Tβ4 and pro-
filin bind to β- or γ-actin with higher affinity than to α-actin,
emphasizing the value of testing actin ligands with specific actin
isoforms. These reagents will make specific isoforms of actin
more accessible for future studies of actin regulation.
cytoplasmic actin | non-muscle actin | β-actin | γ-actin
Correspondence: ridillaj@upstate.edu & ambergd@upstate.edu
Introduction
The actin cytoskeleton is an essential, highly conserved, and
abundant component of cells. Simple eukaryotes tend to ex-
press a single actin isoform, while humans display tissue and
cell specific expression patterns. Although closely related,
actin isoforms can subtly differ in biochemical properties re-
lated to polymer formation, nucleotide hydrolysis and ex-
change, and interactions with one or more essential regula-
tory proteins (Allen et al., 1996; Moradi et al., 2017; Namba
et al., 1992; Perrin and Ervasti, 2010). Humans express six
actin isoforms: α1, α2, α-cardiac, and γ2, which occur pre-
dominantly in skeletal, cardiac, and smooth muscle cells, and
β and γ1 (γ henceforth) found predominantly in non-muscle
cells and considered cytoplasmic isoforms of actin. Human
β- and γ-actin are structurally divergent (Arora et al., 2023),
yet differ by only four amino acids located within their first
ten amino acids (Figure 1A). Various muscle tissues are the
most common sources of actin for use in biochemical studies.
Actin purified in this manner is present as a mixture of muscle
isoforms and minor amounts of β or γ-actin. Characterizing
specific isoforms of actin has been limited by accessibility
and is critically important for understanding mechanisms of
disease (Parker et al., 2020). Thus, it has been challenging to
discern the biochemical properties of either β- and γ-actin in
biochemical assays.
Producing recombinant human actin outside of eukary-
otic cells is difficult due to the complex network of chaper-
ones needed to properly fold actin, however several systems
have been developed (Geissler et al., 1998; Grantham, 2020;
Millán-Zambrano and Chávez, 2014; Schafer et al., 1998;
Valpuesta et al., 2002). The pCold system permits the bac-
terial synthesis of recombinant tagged β-actin (Tamura et al.,
2011). Actin isoforms expressed and purified from popular
eukaryotic systems produce relatively large quantities of bio-
chemically active actin (Bergeron et al., 2010; Bookwalter
and Trybus, 2006; Ohki et al., 2009; Rutkevich et al., 2006;
Yamashiro et al., 2014). While these and related purification
methods have been adopted for various studies of normal and
mutant versions of actin, preparations are often contaminated
with low amounts (5-15%) of host actin (Hundt et al., 2014;
Müller et al., 2012; Müller et al., 2013; von der Ecken et al.,
2016). Other systems use a combination of affinity tags and a
direct fusion to the actin monomer binding protein thymosin-
β4 (Tβ4) to prevent the spurious polymerization or aggrega-
tion of recombinant actin and to facilitate isoform specific
purification (A et al., 2020; Hatano et al., 2018; Hatano et
al., 2020; Kijima et al., 2016; Lu et al., 2015; Noguchi et al.,
2007). With the addition of NAA80 or SETD3 modifications
this approach permits the isolation of actin in specific post
translationally modified states, including N-acetylation, N-
arginylation, and methylation of a conserved histidine residue
(Arora et al., 2023; Hatano et al., 2018; Hatano et al., 2020).
Additional approaches even permit the close preservation of
native secondary modifications (Ceron et al., 2022). While
each system requires post-purification processing, they pro-
vide a source of non-muscle, normal or mutant (including
non-polymerizable), actin isoforms.
We have taken a different approach to generate pure hu-
man β- or γ-actin, engineering the yeast Saccharomyces cere-
visiae to produce either isoform as their only source of actin.
This technique was first pioneered to purify chicken β-actin
from yeast, relying on hydroxylapatite chromatography to
separate host and recombinant actin (Karlsson, 1988). A
follow-up study showed that yeast could survive, albeit not
well, with this as their sole actin source, although β-actin
was not purified from these yeast strains (Karlsson et al.,
1991). We have taken codon-optimized genes for human β-
and γ -actin and expressed them in yeast lacking the resident
actin gene, ACT1. The resulting strains grow considerably
Haarer et al. | bioRχiv | April 3, 2023 | 1–20
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted April 3, 2023. ; https://doi.org/10.1101/2022.08.17.504301doi: bioRxiv preprint