New robot gives LBL Human Genome Center the edge

March 13, 1992

By Lynn Yarris,

LBL's Human Genome Center has become one of the major distributors of DNA clone libraries to research organizations across the nation thanks to a high-speed automated colony picker.

Designed by the Genome Center's instrumentation group under the leadership of physicist Joe Jaklevic, this machine can pick and array colonies of yeast or bacteria at the rate of nearly one colony per second, or more than 2,000 an hour, which is about what an experienced human can total after an eight-hour work day.

Instrumentation group members working on this project include -- in addition to Jaklevic -- Tony Hansen, Marty Pollard, Bill Searles, and Ed Theil. Group member Don Uber worked on an earlier version, which was a general purpose robot that had been adapted to the task of colony picking.

Human chromosomes are far too large to be handled intact by any known biological procedure and therefore must be broken into smaller pieces. These fragments of human DNA are then cloned in colonies of yeast or bacteria cells which are carefully arranged in ordered libraries so that geneticists can later determine their original chromosomal position. Assembling such libraries, which may consist of 50,000 or more individual fragments (each growing in its own colony) requires that cells of yeast or bacteria growing at random in Petri dishes be picked up and transferred to a growth medium in microtiter plate wells. The task is traditionally done by hand with a toothpick.

Since any mistake in the order of this transfer can create problems that may haunt researchers for years, many top scientists insist on picking their own colonies rather than delegate the task to technicians or students. Nonetheless, after several hours of tedium, the potential for error by the most patient of humans becomes high.

In addition to eliminating errors that crop up because hands and eyes become tired, the automated colony picker also frees up the time of scientists for more creative endeavors. This makes colony-picking a far more cost-effective and less time-consuming procedure.

The automated colony picker starts with two tables capable of computer-controlled movement along the x and y coordinates of a grid. The tables are positioned underneath a pair of plungers with a rotating carousel wheel in between that is fitted with 24 reusable needles.

To operate the system, a Petri dish containing colonies of yeast or bacteria is placed on one table, and a 96-well microtiter plate is placed on the other. Guided by a digital image that shows the location of colonies in the dish, the first table moves a colony directly underneath a plunger which at the same time is being armed with a needle from the carousel. On a signal from a computer, the plunger dips the needle into the colony and immediately retracts it with "picked" cells. The carousel then rotates the needle over to the second plunger and the table bearing the microtiter plate moves until the correct well is directly underneath it. The computer activates the second plunger, dipping the needle into the growth medium where some of the cells will be deposited. After retraction from the growth medium, the needle is rotated by the carousel through a tank filled with sterilizing liquid and made ready for reuse.

The system is designed so that while one needle is picking, another is depositing, and another is being sterilized.

"In our first large-scale test, we picked and arrayed a library of nearly 11,000 colonies in one ten hour run," Jaklevic says.

When fully operational and interacting with other components of the automation system being designed for LBL's Human Genome Center, the instrumentation group expects their colony picker to be able to pick and array a million or more colonies a year.

Says Hansen: "At this time, the machine is functioning as a prototype, requiring operator attention and manual loading and unloading. In its completed form, the unit will be designed to operate in conjunction with a general purpose laboratory robot as well as autonomously on a stand-alone, single-use basis. In this latter mode, a person will bring a plate of colonies, place it in the machine, and have the array of picked colonies delivered after a minute or two."

The automated colony picker is now located in Bldg. 70A. Plans call for it to be delivered this spring to the Human Genome Center in Bldg. 74, where it will join the Hewlett-Packard microassay system robot that has been adapted and is now being used to fill microtiter plates with liquid growth medium.