Algae Genetic Screening

Use of the Singer ROTOR HDA for Chlamydomonas Reinhardtii Cell Culture

Fred Cross

The Rockefeller University, New York, NY 10065



The ROTOR HDA was initially designed primarily for use in budding yeast (Saccharomyces cerevisiae) genetics. We are carrying out broad genetic screens in the green alga Chlamydomonas reinhardtii, for which the ROTOR HDA technology would be very useful. However, Chlamydomonas differs from budding yeast in cell size, growth rate, colony morphology and adhesiveness, and light responsiveness. Therefore, it was important to test specifically if the ROTOR HDA technology would work with Chlamydomonas.
In this Applications Note I address the following issues:

  1. Are cells reliably picked up from liquid or solid medium onto the RePads replicaplating pads for the ROTOR HDA); are the cell numbers/density reasonably
    reproducible between pins, between pads and between plates?
    Reproducible pickup and dispensing in terms of number of cells will depend on surface properties of the cells; how clumpy or sticky they are, how much they stick to (or are repelled by) the transferring surface, etc. Chlamydomonas is different from yeast in these properties. For example, in replica-plating using velvets (standard method for yeast), Chlamydomonas colonies have a strong tendency to either stay entirely on the plate or transfer entirely to the velvet.
  2. Do cells retain high viability through this procedure?
    Independent of number of cells transferred, it is obviously desirable for high viability to be retained in the transferred cells. It is clear from our initial work that in many respects Chlamydomonas is more susceptible to damage during some operations than yeast or E. coli. For example, when picked from agar onto a glass or stainless steel needle, Chlamydomonas will lose viability almost completely after ~40 seconds in the air.
  3. Is there significant cross-contamination between pins or across the source or target plates?
    Depending again on properties of the organism, it could easily be imagined that a spray of viable cells could be spread by ROTOR HDA replicating operations, resulting in cross-contamination and ruined experiments.
  4. What density of colonies is the maximum that can reliably be plated and distinguished using the ROTOR HDA?
    The maximum density that the colonies can be plated is dependent on the size and the growth rate of the organism, and will affect the throughput of the screens that can be performed on the organism.



I used Singer PlusPlates™, filled with 50 ml of TAP agar (standard Chlamydomonas medium) supplemented with 50 μg/ml ampicillin to suppress bacterial contamination. Because Chlamydomonas is highly motile in liquid medium and also highly lightresponsive, we removed the interior lights from the ROTOR HDA, and kept the cover closed during operations to the extent possible. Without this precaution, the algae strongly aggregated in different regions of the source plate, obviously confounding reproducibility of spotting across the plate.

Using an open plate filled with 25 ml of liquid TAP medium, containing wet Chlamydomonas (cc-124 background) at ~10^6 cells/ml, we made liquid-to-agar transfers using the long-pin 384 RePad™, in a 384 -> 1536 arraying pattern (revisiting the source between pinnings). Plates were incubated for 3 days at 33°C under strong illumination (Figure 1A), and were examined microscopically at intervals.

Liquid-to-agar transfer. ROTOR HDA application, Algae Genetic Screening

Figure 1: Liquid-to-agar transfer.

(A) Photograph of a plate showing 1536 60 nl drops of Chlamydomonas suspension spotted with 384-long RePad™ after 2 days incubation. The photograph shows uniformity across spots.
B) Micrograph of region in 1 of 1536 drops after plating (0 hr.)
at high- (top) and low- (bottom) magnification.
(C) Same region after 18 hours incubation (right) at high- (top) and low- (bottom) magnification. Viability is >95% (all cells form microcolonies).


Microscopic examination showed that the pinned spots were approximately 1.5mm in diameter, with fairly random distribution of cells across the spot (Figure 1B). Microscopic counts of cells per spot plated indicated 62 +/- 11 cells per spot (mean +/- sd; n=56). The error is greater than predicted for a Poisson distribution, implying variability in actual volume transferred; a rough calculation suggests a volume distribution of ~60 +/- 10 nl, with occasional outliers.

High viability is evident since essentially every cell plated formed a microcolony by 18 hrs. (Figure 1C). A macroscopic image of the plate demonstrates reproducibility across the pinned area. Good reproducibility between pinned plates was also observed (data not shown).

At this pinning density there was clear spatial separation between spots, with no evidence of contaminating colonies in between.

In other experiments we have observed sporadic occurrences of ‘heavy’ pinning events, where a volume of ~120 nl is transferred. When this occurs, it is plate-wide, but specific to an individual pinning event. Because of the sporadic nature of these occurrences we have been unable to determine how they come about. It is not reproducible for a given RePad™, source or target plate.

The ROTOR HDA requires ~5-10 sec between pickup and plating. To determine viability of Chlamydomonas held in air on the RePad™, I paused operation after pickup and before plating for an additional 0, 10, 20, and 40 seconds. The number of cells transferred and their viability on incubation remained essentially unchanged.

I tested liquid-to-agar transfer from a Singer plate containing 50 ml of cell suspension at the same density rather than 25. Results were essentially identical, indicating that the effective drop size is largely independent of the depth of liquid into which the pin is inserted.

Fred Cross Rockefeller University - Chlamydomonas reinhardtii cell culture with the ROTOR HDA

See Fred trialling ultra high-density algae pinning using the ROTOR HDA at Rockefeller University.



A lawn of cc-124 background Chlamydomonas on a Singer PlusPlate™ containing TAP agar was used as a source. A 384 long-pin RePad™ was used to pick up and transfer cells to a fresh plate, pinning 16 consecutive spots without a revisit to the source. This results in 6144 spots, in sets of 16 that progressively dilute the inoculum. These plates were examined microscopically, and incubated to grow up colonies.

Figure 2 shows images of the spots. Spot diameter is ~1 mm. The sequential pinnings result in progressive diminution of the number of cells transferred, although this effect is quantitatively irregular; thus, this method can be used to reduce inoculum density down to isolated single cells, although the dilution factor per strike is variable. Cell viability upon agar-to-agar pinning is generally high, probably comparable to liquid-to-agar pinning.

In general, both individual cells and resulting microcolonies and colonies stay within spatial borders dictated by the pin geometry, because Chlamydomonas is non-motile on plates, and the procedures are accurate with little effective spray or aerosol.
Therefore, these procedures are suitable for effective parallel transfer of large numbers of individual Chlamydomonas cultures, in a very space- and materials-efficient format.

Agar-to-agar transfer. ROTOR HDA application, Algae Genetic Screening


ROTOR Synthetic Genetic Array Robot product link

The ROTOR is the fastest and most powerful colony manipulation robot in the world. It is essential for any large-scale Algae studies.


Figure 2: Agar-to-agar transfer.

A lawn of Chlamydomonas was pinned with a 384-long RePad, and cells were pinned to a fresh plate 16 times without a revisit.
(A) Micrograph (montage) of cells immediately after pinning; progressive dilution can be seen.
(B) A region of the plate after 2 days incubation. The effects of dilution are detectable as differential amounts of growth.




We are interested in the maximum density at which parallel cultures of Chlamydomonas can be maintained using the ROTOR HDA technology. A Chlamydomonas lawn was used as a source for pinning with the 6144-density RePad™. This resulted in quite even transfer of small inocula onto the target plate, with no evidence of contamination between spots (Figure 3A). These 6144 individual cultures grew well into well-separated 0.5 mm diameter colonies (Figure 3B). The grown-up 6144 array was used as a source for a subsequent pinning 1 day later, and excellent alignment of the grown array and the subsequent pin-strikes was observed (Figure 3C).

Overall, the 6144-density is usable with Chlamydomonas, and promising for future work; probably the main modification compared to work at 96 to 1536-density is the need for flat plates (without the usual meniscus at the edges and corners, and also lacking any other irregularity in the surface) since at 6144-density this can seriously perturb transfer either across a region, or in individual spots.

Figure 3D shows the result of 16-fold pinning (with revisit) of a 384-array of randomly selected Chlamydomonas mutants (UV mutagenesis). Strong differences in growth rate are obvious among the mutants, and these differences are highly reproducible across the 16 replicates. This suggests that 6144-density might be usable with a highly complex assembly of different genotypes, on one plate.

6144-density is, in my opinion, completely unworkable with standard microbiological methodologies (velvets, toothpicks, etc) because the spatial resolution of these methods (even for expert workers) is seldom at the sub-mm resolution needed at this high density. Therefore, the ability to work at this density is a unique advantage of the

Agar-to-agar transfer at 6144 density. ROTOR HDA application, Algae Genetic Screening

Figure 3: Agar-to-agar transfer at 6144 density.

A lawn of Chlamydomonas was pinned to a fresh plate using the 6144 RePad.
(A) An area of the plate after 2 days incubation. Growth is even for all inocula across the field.
(B) Micrograph of higher regular resulting microcolonies obtained after growth.
(C) Reproducibility of pin-strikes. A 6144 array of colonies was used as a source for a second pinning with a 6144 RePad. The second pin-strike was well within centerto-
center spacing of the first one, carried out with a different pad 1 day previously.
(D) A collection of 384 randomly UV-mutagenized Chlamydomonas clones was replicated 16x (with revisit) and grown up for 2 days. Note high reproducibility between the 16 replicates of growth rate variability associated with each mutant.



Chlamydomonas reinhardtii is fully compatible with the Singer ROTOR HDA for liquid-to-
agar, agar-to-agar or agar-to-liquid transfers. In general, viability and quantitative reproducibility of transferred cells appears to be high. The 6144 array density is likely workable, although demanding in terms of agar plate quality, and could be an efficient format for high-throughput genetic assays. It can likely be integrated effectively with direct microscopic examination for phenotypic classification, thus in some systems eliminating the need for any macroscopic testing or culturing.





See Fred trialling ultra high-density algae pinning at Rockefeller University





Algae Genetic Screening Applications

High-Throughput Genotyping of Green Algal Mutants Reveals Random Distribution of Mutagenic Insertion Sites and Endonucleolytic Cleavage of Transforming DNA
Abstract (Zhang R, Patena W, Armbruster U, Gang SS, Blum SR, Jonikas MC)

A high-throughput genetic screening platform in a single-celled photosynthetic eukaryote would be a transformative addition to the plant biology toolbox. Here, we present ChlaMmeSeq (Chlamydomonas MmeI-based insertion site Sequencing), a tool for simultaneous mapping of tens of thousands of mutagenic insertion sites in the eukaryotic unicellular green alga Chlamydomonas reinhardtii. We first validated ChlaMmeSeq by in-depth characterization of individual insertion sites. We then applied ChlaMmeSeq to a mutant pool and mapped 11,478 insertions, covering 39% of annotated protein coding genes. We observe that insertions are distributed in a manner largely indistinguishable from random, indicating that mutantsin nearly all genes can be obtained efficiently. The data reveal that sequence-specific endonucleolytic activities cleave the transforming DNA and allow us to propose a simple model to explain the origin of the poorly understood exogenous sequences that sometimes surround insertion sites. ChlaMmeSeq is quantitatively reproducible, enabling its use for pooled enrichment screens and for the generation of indexed mutant libraries. Additionally, ChlaMmeSeq allows genotyping of hits from Chlamydomonas screens on an unprecedented scale, opening the door to comprehensive identification of genes with roles in photosynthesis, algal lipid metabolism, the algal carbon-concentrating mechanism, phototaxis, the biogenesis and function of cilia, and other processes for which C. reinhardtii is a leading model system.