TECHNICAL NOTE

20
Q Q Q Q Q Q Q Q
52

Tips for Purchasing Research Fermentors and Bioreactors

A Practical Guide for Researchers

Julia Cino and Stanley Frey In this three-part series, the authors provide a practical guide for purchasing research fermentors and bioreactors. Part 1 describes five ways to avoid making costly mistakes. Look for the remaining 15 tips in upcoming issues of BioPharm.

Julia Cino is product manager, and corresponding author Stanley Frey is director of advertising at New Brunswick Scientific Co., Inc., P.O. Box 4005, 44 Talmadge Road, Edison, NJ, 08818-4005, (800) 631-5417, fax (908) 287-4222, bioinfo@nbsc.com

Choosing laboratory fermentors and cell culture bioreactors can be a complex matter. It usually requires experience and technical expertise to fully comprehend the specifications and engineering nuances of various competitive fermentors. A major problem lies in the lack of specific information required to make intelligent decisions. For various reasons,

manufacturers often omit critical product specifications and performance data from product literature or formal quotations. Some may not have the research facilities or technical staff to conduct the required performance tests, and others may be reluctant to reveal unfavorable data that might impede sales. Whatever the reason, this omission can leave a gaping hole in your information-gathering process. In purchasing fermentors and reactors you must address vessel design and sterilization problems. Consider questions of mass transfer capability, bacterial contamination, and FDA validation requirements. Knowing which design features to look for and the right questions to ask can take the mystery out of the decisionmaking process and keep you from making costly mistakes. (See Ask the Right Questions box.) With so much at stake, it is imperative that you ask critical questions and receive meaningful answers. Asking the right questions can help you cut through the veil of confusion and correctly assess the advantages and disadvantages of the products available. Equipment manufacturers can have significantly different design and fabrication standards. Users needs also differ. Some may require only a simple fermentor, whereas others, because of budgetary constraints, may invest in only a bare bones culture vessel. For users with more sophisticated research or specific

ASK THE RIGHT QUESTIONS

Which types of glass vessels will shorten sterilization and cool-down cycles rather than prolong them? Which glass vessels are more vulnerable to breakage? Can glass vessels be safely steam-sterilized in place? Are threaded ports in the fermentor headplate designed to completely eliminate the risk of contamination? Which stainless steel surface treatments are FDA validatable? Are all mirror-finished stainless steel vessels free of unsanitary microscopic crevices? How can you be sure the reactor is capable of rapid heat-up and cool-down? Are filters supplied for the fermentor but completely overlooked for addition vessels and accessory ports?
BioPharm
SEPTEMBER 1996

Q Q Q Q Q Q Q

Will prefiltration and regulation of air and water services be necessary for dependable operation? Does the culture system have the capacity to deliver a sufficient supply of oxygen to meet the demand of highly aerobic organisms? Will you be able to perform temperature induction studies with your new equipment? Are analog or digital systems more reliable? Will your new fermentor be able to communicate with your existing recorders and controllers? Can the system maintain the desired temperature and rate of agitation while overcoming the added heat load and viscosity of densely growing organisms? Will you receive start-up assistance and after-sale support for your bioprocessing software?

process needs, a simplified system may not be enough. Whatever your needs, weve identifed 20 ways to avoid making common and costly mistakes when purchasing fermentors and bioreactors. Here are five of them. 20 TIPS FOR PURCHASING

1 Choose the Right Vessel

for Your Process
Temperature control can sometimes go awry over the course of a fermentation in which high concentrations of biomass are produced, especially when glass-jacketed reactors are selected as the culture vessel. Glass is a poor conductor of heat and has approximately one-fortieth the heat-transfer capability of stainless steel (1). The exothermic reaction of cultures growing at densities of 50 to 100 grams per liter can overwhelm temperature control in glassjacketed vessels. At ambient temperatures, circulating cold water in the jacket is not always enough to compensate for the heat output of high-density cell growth. During exponential growth, insufficient cooling capacity causes operating temperatures to rise significantly above the set point. This problem also surfaces during the sterilization cycle because glass limits the heat-exchange capacity of the vessel. The gravity of this problem was clearly demonstrated during a fermentation workshop in which three different types of laboratory fermentors were simultaneously autoclaved with the same medium in the same autoclave. Fermentor vessel Type A was a conventional flat-bottom glass jar with a stainless steel headplate (Figure 1). Vessel Type B was a cylindrical glass tube mounted between a stainless steel headplate and a dished-bottom jacket of stainless steel. Vessel Type C was an all-glass jacketed

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Choose the right vessel for your process. Know the risks of steam-in-place glass fermentors. Avoid mishandling glass fermentors during sterilization. Beware of unsanitary threads and fittings. Know which surface finishes and treatments are FDA validatable. Demand proof of performance. Avoid glass condensers. Know whether manfacturer pays for prefilter kits for air, water, and steam. Is your system designed for temperature induction? Specify unbreakable side-wall vessel ports. Compare analog with digital controllers. Make sure the new fermentation system communicates with your existing instrumentation. Make sure you get all the parts and pieces you need. Know what instruments are included in the manufacturers quote. Determine your long-term requirements for feeding additives. Ask the manufacturer about its FDA validation package. Know who will service your system and where. Use your PC to its fullest extent. Ask about start-up assistance and service agreements. Make sure your system can produce the required dissolved oxygen concentration.
BioPharm
SEPTEMBER 1996

vessel beneath a stainless steel headplate. Upon removal of the fermentors from the autoclave, the medium in the first two reactors appeared normal. However, investigators were uncertain about the sterility of the medium in the glass-jacketed vessel. Acting as a thermal insulator, the air space in the glass-jacketed fermentor impedes the flow of heat to the medium, preventing it from reaching the required temperature in the expected time. According to heat flow studies (conducted by New Brunswick principal investigator Y. Chen at Rutgers University in 1996) with stirred jar fermentors in an AMSCO Scientific (Apex, NC) Model 72A wall-mounted autoclave, the stainless steeljacketed fermentor reached sterilization temperature in less than half the time of the glass-jacketed fermentor. With thermocouples immersed in the culture medium of each vessel, the medium in a 5-L Type B fermentor reached 121 C in 35 minutes, compared with the glass-jacketed fermentor, which was unable to reach 121 C even after 90 minutes (Table 1). A similar problem occurs during the cooldown cycle. Vessel Type C requires a much longer cooling time because of the slower transfer of heat through the air space in the glass jacket. To remedy those problems, some researchers fill the glass jacket with water to speed up heat transfer. Unfortunately, when the jacket of a 5-L fermentor is filled with 11.4 L of water and then autoclaved, tests show that it takes 55 minutes to attain a

2 Know the Risks of Steam-

in-Place Glass Fermentors

Can glass jar fermentors be safely steamsterilized in place? As far as can be determined, sterilization in place of glass fermentors and bioreactors involves some risk if you select the wrong system. Laboratory culture vessels can withstand repeated sterilization in an autoclave because vessels are vented, which equalizes the pressure inside and outside the vessel. But if the pressure inside the vessel becomes greater than the pressure outside (as occurs when pressurized steam or gas is

fed directly into the glass jar), the glass whether borosilicate or Pyrex can crack or burst. Such danger applies primarily to glass jar vessels (Type A), which are vulnerable to breakage, especially where the base is joined to the tubular walls. On the other hand, Type B vessels are less susceptible to breakage because the number of stress points is significantly reduced by the tubular design of the glass vessel (Figure 2). Do not be misguided by a pressure gauge or pressure-relief valve mounted in the vessel headplate. It offers no protection against the stress placed on glass that can burst at pressures below 5 psig. For this reason it is unsafe to pressurize a glass

54

sterilization temperature of 121 C a heat-up time that can result in excessive heat exposure and high turbidity for some media. Avoid a glass jacket that is permanently closed, because the jacket interior cannot be cleaned of debris and condensation. Look
Table 1. Sterilization of bioreactors.
Vessel Size (liters)

for a jacket that is open and flanged at the bottom where it is sealed against a protective steel baseplate. This adds little or no extra cost to the vessel and provides the convenience of good visibility and protection against glass breakage.

Fermentor Type Cylindrical glass tube with dished stainless steel jacket (Type B) All-glass jacketed vessel (Type C)

Maximum Temperature (C)

Sterilization Time (min)

Jacket Water (liters)

5 7.5 5.0 2.2

121 119 121 121

35 None BP SEPT 96 FREY FIG.1 None 90 Filename: 9FREYF1 55 1.0 55 1.4

Knowing the configuration and surface area of the jacket or cooling coil is insufficient to evaluate the heat transfer efficiency of the reactor. To determine whether a reactor design can meet the heating and cooling requirements for a particular process, investigators must know the heat output of the culture at maximum cell density. Place that responsibility with the vendor by specifying the heat-removal capability of the equipment, expressed in watts per liter. Find out before you buy whether temperature can be maintained with the circulation of city water (at the citys highest lab temperature) or whether the vendor requires an expensive accessory chiller system to meet your lowest temperature requirement.

Unbreakable side ports Glass tube reactor

Glass hose barb (water outlet port) . B . Hemispherical stainless steel jacket Glass fermentor with stainless steel bottom-dished jacket All-glass jacketed fermentor vessel Glass jacket C

Glass hose barb (water inlet port)

Quick-connect stainless steel inlet/outlet ports Stainless steel base plate

Glass jar fermentor

Figure 1. Fermentor vessels. Type A: Flat-bottom glass jar with stainless steel headplate. Type B: Cylindrical glass tube mounted
between stainless steel headplate and dished-bottom jacket and baseplate. Type C: All-glass jacketed vessel mounted beneath stainless steel headplate.

reactor in the open environment of the workplace. At high temperatures and pressures, glass can break at weak stress areas where glass thickness varies. Such differences in thickness can vary as much as 6 mm. Such variations can cause multiple stress points that are vulnerable to damage during sterilization. Stress on the glass is exacerbated by the added tension resulting from different expansion coefficients of metal and glass. Several fermentor manufacturers offer an inexpensive sterilization shroud designed to contain accidental glass breakage during sterilization. Although some shrouds afford

protection against glass breakage, they offer little or no protection against hot medium spills. Unless internal and external pressures are equal during sterilization, and unless the glass vessel is completely protected by a pressurizable steel safety hood, in situ sterilization of glass jar fermentors is not recommended.
Figure 2. Avoid accidental spills in steam-inplace glass jar fermentors by ensuring that your fermentor is protected by a pressurizable dome that is anchored securely over the glass vessel during sterilization. Shown here is a pressurizable dome of stainless steel that clamps securely to the base of the fermentor.

BioPharm

SEPTEMBER 1996

55

3 Avoid Mishandling Glass

Fermentors during Sterilization

Figure 3. Ensure that your steam-in-place

fermentor can be completely sterilized in place. Allow for the simultaneous sterilization of accessories, such as addition vessels, feed lines, and filters as illustrated.

It is commonly known that special care must be exercised in autoclaving an all-glass vessel or reactor, but accidental breakage during sterilization and handling still occurs. A mistake of this type can be costly

from $450 to more than $1,000 for a 5-L replacement vessel, depending on its country of origin. A frequent mistake is to remove the vessel from the autoclave too early. Make sure that the autoclave is properly vented during sterilization and that the vessel is removed only after it has cooled for the length of time specified by the manufacturer. Unless your lab personnel are able to carry heavily laden, 10-L (and greater) glass fermentors, avoid purchasing large-volume glass reactors. When a heated reactor is removed from the autoclave, the tensile strength of the glass is significantly weakened and is highly vulnerable to breakage. Many SIP benchtop fermentors, whether made of glass or stainless steel, cannot be sterilized in their entirety. Accessory syringes, connectors, samplers, reservoirs, and transfer tubing for acid, base, and antifoam must be autoclaved separately and then aseptically connected after sterilization. There is little advantage to an in-place sterilization system if an operator must carry the accessories to an autoclave and then sterilize them separately. If an SIP system is complete and your system has sufficient capacity to accommodate the sterilization of accessories, you can avoid time-consuming trips to the autoclave (Figure 3).

5 Know Which Surface

Finishes and Treatments are FDA Validatable

4 Beware of Unsanitary Threads and Fittings

Every opening in the fermentor should be designed and constructed for maximum protection against contamination. If headplate ports and penetrations have internal threads, ensure that they are sealed with O rings so that the threads are not exposed to the process side of the reactor. Threads are difficult to clean and can harbor contaminants in tiny crevices where bacteria are not easily destroyed by sterilization. An average reactor vessel contains at least 10 openings in the headplate, each a potential risk of contamination. Many buyers insist that ports be welded to the headplate wherever possible so that the threads can be located on the outer perimeter of the port and unexposed to the process. Ports with internal threads are cheaper to fabricate but are not universally accepted. Avoid ports with internal threads if your equipment must comply with FDA requirements. Unfortunately, in small-size vessels this construction consumes much needed head space and does not allow for a full range of sanitary welded ports. O rings can prevent almost all potential contamination in laboratory and pilot plant environments. But set screws used to anchor impeller blades to an agitator drive can present some risk if they are not carefully washed and cleaned. Submerged in liquid, these screws are sterilized with the vessel and are not typically exposed to airborne contamination.

To satisfy FDA validation requirements, all internal welds must be ground and polished so that no corners and crevices remain where contaminants could possibly lodge. Such a finish is achieved by first mechanically polishing the interior surface to a 20-micro-inch roughness average (Ra), followed by electropolishing and passivation. (Exterior surfaces need be polished only to a 35-micro-inch Ra to facilitate cleaning.) Electropolishing is an electrolytic dissolution of the metal surface projections, which smooths and brightens the surface. Passivation is a final step that constitutes soaking the fermentor in a caustic solution followed by a nitric acid bath that cleans the metal and creates a chemically inactive surface that is highly resistant to corrosion. All that glitters is not gold. So it is with process vessels. A highly polished mirror finish does not necessarily mean that the stainless steel surfaces are smooth and sanitary. Such bright finishes are deceptive when they are applied over poorly ground surfaces to conceal an unsanitary finish. An electropolished finish is more sanitary than mechanical polishing, which can leave microscopic crevices that trap particulate matter. Not all equipment is manufactured to these standards. Some are mechanically polished and then passivated, bypassing the electropolishing step. Most laboratories accept this standard, but to ensure that you get what you want, these requirements should be clearly defined in your specifications.

Looking Ahead

Part 2 of this three-part series continues with more tips for purchasing research fermentors and bioreactors.
References (1) J.M. Coulson and J.F. Richardson, Thermal Conductivities, Chemical Engineering, Vol. 1: Fluid Flow, Heat Transfer, and Mass Transfer, J.R. Backhurst and J.H. Harker, Eds. (Pergamon Press, Oxford, UK, 1977), 174. BP

56

BioPharm

SEPTEMBER 1996

42

BioPharm

SHOWCASE

FEBRUARY 1997

TECHNICAL NOTE

20
1 2 3 4 5

Tips for Purchasing Research Fermentors and Bioreactors

A Practical Guide for Researchers, Part 2

Julia Cino and Stanley Frey This concludes a two-part guide for purchasing research fermentors and bioreactors. Part 1 included tips 1 through 5 and appeared in the September 1996 issue of BioPharm.

TIPS FOR PURCHASING 1 through 5

Choose the right vessel for your process. Know the risks of steam-in-place glass fermentors. Avoid mishandling glass fermentors during sterilization. Beware of unsanitary threads and fittings. Know which surface finishes and treatments are FDA validatable.

Avoid Glass Condensers

6 Demand Proof of Performance

Before you buy a steam-in-place (SIP) fermentor or bioreactor, ascertain how long the system takes to complete sterilization and cool-down cycles. Dont wait until the equipment has been delivered to learn that there is an unexpectedly long delay in reaching sterilization and cool-down BP SEPT 96 temperatures. FREY FIG.2 Filename: 10 FREY F2 Some systems allow users to select heat-up time as well as the sterilization period a feature that is particularly useful in mimicking heat-up times for very large fermentors. To discover bacteriological leaks or performance problems in a fermentation system before shipment, manufacturers of SIP fermentors usually conduct a 48- to 72-hour sterility test in which the heat-up and cool-down temperatures are recorded (Figure 1). If this record is not included in the standard documentation package, ask for it in the purchase order. In addition to revealing sterilization and cooldown times of the vessel contents, records should tell you whether growth temperature can be maintained within the levels of accuracy specified in the manufacturers literature and whether the system is properly engi(a) 125.0 neered for rapid heat-up and cool100.0 down. This information should be 75.0 readily available from the manufac50.0 turer. 25.0 The system should be designed 0.0 for cool-down after sterilization by 00:00 18:00 36:00 54:00 72:00 water circulation. For non-SIP vesTime (hr) sels, manufacturers should provide (b) information about the time required 125.0 to reach sterilization temperature (of 100.0 water) at maximum working 75.0 volume. 50.0 25.0 If conserving water is a consider0.0 ation, the fermentor should be 00:00 00:30 01:00 01:30 02:00 designed for continuous recirculaTime (hr) tion of chilled water. To avoid disposing water down the drain, the Figure 1. Manufacturers 72-hour sterility test of a steam-in-place 20-L benchtop system should be fabricated with a fermentor in which key operating parameters including pH and dissolved oxygen water jacket or a cooling coil in con(DO) are recorded (not shown in temperature profile). Profile (a) plots the heatjunction with an accessory chiller up and cool-down time from ambient to 125 C over a 72-hour period. Profile (b) that recycles coolant back to the displays greater detail in an expanded scale of a two-hour segment of the sterility fermentor. test.
Temperature (C) Temperature (C)

Exhaust gas condensers protrude from the headplate and can be easily broken if made of glass. A stainless steel condenser can be much more expensive, but it is unbreakable and a superior heat exchanger more efficient at condensing the exhaust gases lost in evaporation and returning them to the culture (Figure 2). In addition, avoid the simple cold finger (tube-within-a-tube design). To maximize heat-exchange capability in the condenser, a turbulent flow of gas is required. One way to improve heatexchange efficiency is to pass exhaust gases in a tortuous path around an inner cooling coil for vigorous mixing and BP NOV the enhanced contact with the surfaces of 96 cooling coil.
FREY FIG.4 Filename: 10 FREY F1

Exhaust

Cold water in/out Cooling coil Fermentor cover plate

Fermentor vessel

Figure 2. To avoid increased viscosity of the culture

medium over time, your bioprocessing system should be equipped with an exhaust gas condenser to minimize evaporation loss. Determine whether the device is made of glass or stainless steel and whether it is designed to create a turbulent flow of gas for enhanced heat transfer.

Julia Cino is product manager, and corresponding author Stanley Frey is director of advertising at New Brunswick Scientific Co., Inc., P.O. Box 4005, 44 Talmadge Road, Edison, NJ, 08818-4005, (800) 631-5417, fax (908) 287-4222, email (bioinfo@nbsc.com).

44

BioPharm

SHOWCASE

FEBRUARY 1997

BP Oct 96 Frey Fig. 3 Filename: 10FreyF3

8 Know Whether Manufacturer Pays for Prefilter Kits for Air, Water, and Steam
Buyers often forget to ask about installation and connection of equipment to air, water, and steam facilities. Remember to allocate funds for prefiltering and regulating those services if necessary (Figure 3). If particulate matter is not removed from water, air, and steam supplies, it can accumulate as sediment in solenoid valves. There it can cause the valve plunger to freeze in place or prevent the valve seat from properly sealing against the plunger. Such deposits can strike a damaging blow to temperature control systems when cold or warm water fails to circulate or shut off on demand. In addition, water and air supply lines may be highly pressurized and can rupture filters if not controlled by in-line pressure regulators. Know which manufacturers include these preassembled piping assemblies in their price and which do not. Some state and local licensing laws prohibit out-of-state manufacturers from making the necessary piping connections in a research facility, so some manufacturers are unconcerned with the connection of equipment to utilities. A single prefiltration and regulation hook-up can often be adapted for more than one bioprocessing system. Before these components are installed, ask your building superintendent whether the piping is needed. Then make sure your water, steam, and air services are adequate for the system you intend to purchase.
Steam prefilter assembly

11

Compare Analog with Digital Controllers

Air prefilter assembly

Dont be misled by analog equipment of the past dressed up with modern digital displays for temperature, speed, and other parameters. With analog instruments, zero and span are interactive and therefore difficult to calibrate accurately. Accordingly, they may display erroneous readings for values measured over an entire control span. Avoid analog systems with signals that are fed to a signal processor located centrally in a facility. With analog systems, a bundle of 20 or more wires can extend over long distances where signals can be easily corrupted by spurious impulses or electronic noise generated in the facility. Digital systems are more reliable with no more than four wires required to transmit the data contained on all loops combined. If your system is digitally controlled, signals can be optically isolated and protected against ground loops. Signal conditioning takes place on a signal-conditioning card or mother board mounted at the console where all analog signals are converted to digital.

Water prefilter assembly

Figure 3. Because piping connections to utilities that regulate and prefilter steam, air, and water are not always required, these connections can be overlooked by the seller.

9 Is Your System Designed for

Temperature Induction?
Temperature induction can be an effective tool for expressing many important proteins. A temperature shift can be performed manually or automatically with computer software that allows you to establish a time-driven table of temperature set points. But if the system hardware is not designed for rapid heating, it may be difficult to control. This is purely a function of the heat-exchange system, which should provide a sufficiently large surface area to heat the maximum volume of culture at a rate of not less than 1 C per minute over a range of approximately 3242 C (1). Temperature shifting has been used with many cell lines to increase product yield. A hot finger is an inexpensive approach to the problem but is not always the best choice, especially for propagating highly aerobic organisms or in those applications where temperature of the water supply is a desired temperature for the process.

12

Make Sure the New Fermentation System Communicates with Your Existing Instrumentation

Dont paint yourself into a corner by ordering a new analog fermentor that is incompatible with the digital recorders and controllers currently in your laboratory. Dont wait until its too late to find out that the new bioreactor you purchased cannot communicate with your PC. Coupling an analog-todigital and digital-to-analog converter to the system allows all types of equipment to interface with each other (Figure 4). This electronic converter should be capable of changing digital signals into 4-20 mA signals used by analog equipment. If the converter can change 4-20 mA signals into an RS-232 or RS-422 format, then the equipment will be compatible with computer software for data logging and programmed control.

Figure 4. Your new equipment should be compatible with the equipment currently in your facility. If its not, a universal converter can be used to talk to both analog and digital instrumentation.

13

Make Sure You Get All the Parts and Pieces You Need
computer software, the total volume of all additives can be calculated for you. Just punch in the volume per dose at the inception to determine the total volume of any supplement at any time during the process. Such on-line measurements can help to instantly evaluate the influence of additives on productivity. Ask the seller whether the foam sensor can be used interchangeably to detect foam and liquid levels. If the metering pump can be electronically assigned and programmed for different functions, the added flexibility can sometimes be useful for liquid level control when it is necessary to add supplements to the culture vessel. If you order an SIP system, ensure that it meets your requirements for the aseptic addition of biological materials. If your process calls for repetitive sampling and addition of reagents, you may require resterilizable addition and sampling ports that allow you to steam the inlet and outlet lines before and after each operation.

10 Specify Unbreakable Side-Wall

Vessel Ports
With laboratory bioprocessing equipment, avoid purchasing glass vessels with breakable glass process ports (serrated hose connections) that project from side walls. Broken glass connectors are inevitable and usually necessitate replacing the entire vessel. Polymeric connectors are autoclavable and seldom, if ever, break. Ensure that connectors are removable and designed to be replaced or plugged up.

Most automatic addition systems contain a dozen different parts including the control module. Check your foam control system, for example. Make sure yours is equipped with more than a sensor, a controller, and an addition pump. Does the reactor have an antifoam addition line preinstalled in the headplate, or must you remember to order it separately? Is the system available with a sterilizable antifoam reservoir with a transfer line, tubing, and a filtered exhaust outlet? Some scientists object to hunting down all the bits and pieces needed to assemble a complete system. You can spend thousands of dollars for a complete bioprocessing system. You shouldnt have to rummage through your supply closets and benches for $20 worth of attachments. Depending on the price, some equipment can total the number of antifoam addition cycles, allowing you to calculate the cumulative volume of chemical defoamer and other supplements that may be added to the culture. With the appropriate

46

BioPharm

SHOWCASE

FEBRUARY 1997

14 Know What Instruments Are Included in the Manufacturers Quote

Some manufacturers completely overlook standard components generally supplied with culture equipment. To keep the initial price of the equipment low, manufacturers may deliberately omit basic components such as air flow meters, spargers, inlet and exhaust gas filters, exhaust condensers, addition ports, and pumps for acid, base, and antifoam. Know what the fermentation system can and cannot do. The manufacturer should provide a list of recommended spare parts that includes items such as O rings, inoculation septa, air and exhaust filters, fuses, drive shaft seals, DO membranes, sensors, and electrolytes. If your fermentor is to be used in a cGMP facility, the manufacturer should provide a list of disposable items that are periodically replaced. FDA requires that all fermentor parts be thoroughly cleaned between fermentations to ensure that components are completely free of residual organisms and contaminants, particularly when a cell line is to be changed from one run to the next (2). To comply with FDA standards, researchers in cGMP facilities have adopted the policy of changing all rubber parts on the interior of the reactor (O rings, washers, and gaskets, for example) that can harbor traces of viable organisms and contaminants, even after steam-sterilization. Most bioengineers will replace all disposable components rather than spend time validating the cleaning procedures employed to guarantee the total destruction of organisms. Astute buyers will ask that the equipment quotation contain an itemized list of the components supplied with the reactor system as well as the cost of replacement parts. Because of the vast price differential for circuit boards and glass vessels, for example, many hundreds of dollars can be saved on spare parts, depending on the manufacturer and its country of origin.

15 Determine Your Long-Term

Requirements for Feeding Additives
Dont wait until its too late to discover that the system you ordered can feed only one additive at a time to the reactor. With digital systems you may need a delivery system that can be programmed and assigned to various functions. By keying in the desired information, pumps can be assigned to add acid, base, antifoam, and nutrient. If for any reason one pump is taken out of the process loop, the pump operations may be reassigned electronically to provide the critical control functions. More than one additive can be metered to the culture by assigning more than one pump to the addition function. The flow can be varied without changing tubing size if pumping time is individually adjustable (Figure 5).

16 Make Sure Your System Can Produce the Required DO Concentration

All bioprocessing equipment should have an oxygen transfer rating (OTR) that measures the oxygen transfer efficiency of the culture vessel. OTR is measured in millimoles of oxygen per liter per hour and is frequently determined with a sulfite oxidation test by the manufacturer for each size and type of vessel. This rating can be used as a guideline to ensure that any bioprocessing system can meet the oxygen uptake rate of BP Sept For eukaryotes and prokaryotes.96 cultivation of most convenFrey Fig. 9 tional organisms, an OTR9FreyF9 of 350 mM O2/L/hour Filename: efficiency can be used to support cell densities 100 g/L (dry weight). Unfortunately, this rating may not be very helpful in meeting the dissolved oxygen demands of highly aerobic organisms currently used in bioprocesses. Many scientists find that the host expression system they are using is oxygen-limiting even though fermentor air flow and impeller speed are set at maximum levels. In the growth of recombinant Pichia pastoris, an expression system currently used for the production of many proteins, the yeast cells do not produce high levels of protein, and investigators can do little to meet the high oxygen demand of the organism. In standard systems, only when a system is added to automatically supplement pure oxygen can they produce cell densities 150 g/L (dry weight) and high protein yields. If your culture system cannot regulate the mixing of air and pure oxygen, the manufacturer might consider furnishing the equivalent instrumentation as a plug-in module (Figure 6). Built-in gas mixing instrumentation can do more than supplement the air stream with oxygen. These systems can also be used with gases such as nitrogen to establish a highly controlled gaseous environment for microaerophiles. Or they can be used to adjust pH with the addition of CO2 instead of acid. Adding acid to cell cultures can cause localized damage to mammalian cells growing at slow stirring rates. Researchers who require special gaseous environments can also use the gas-mixing capabilities of a fermentor to proportion a variety of compatible gases.

Figure 5. Digitally controlled fermentors allow pumps to

be electronically assigned various functions, such as the addition of acid, base, antifoam, and nutrient.

17 Ask the Manufacturer About Its

FDA Validation Package
If FDA is to validate your process, choose a manufacturer whose bioprocessing systems come with a complete package of documentation support services for FDA validation. The validation package documents a systems components, materials of construction, testing program of operational checks, sterility testing, and calibration of process loops. This can save considerable time and effort.

Mode 1

Mode 2

Mode 3

Mode 4

Agitation first

Agitation only

O2 Enrichment only

Then O2 Enrichment

Agitation and O2 Enrichment

Figure 6. Make sure that your new bioprocessing system can meet the O2 uptake rate of cultures producing abnormally high cell and protein yields. A system worth considering is a four-way dissolved oxygen controller. If agitation alone is incapable of meeting demand, a second mode of O2 enrichment may be employed. In a third mode, increased agitation may be used, followed by O2 enrichment. As a fourth alternative, both agitation and O2 enrichment can be used simultaneously to optimize DO.

48

BioPharm

SHOWCASE

FEBRUARY 1997

18 Use Your PC to Its Fullest Extent

Of all the mistakes made in purchasing bioprocessing equipment, underusing the computer that controls the process may be the most serious. Most bioengineers using PCs in fermentation and cell culture use their computers primarily as data loggers. Analysis of computer software applications in the bioprocessing field reveals that only 15% of researchers use computers as process development tools. Paradoxically, the computer technology that helped us discover a path to the planets in space exploration is used merely as a recorder in the discovery of biological products. There is some economic justification for this. A single PC with specialized software can monitor eight processes simultaneously at a cost of approximately $800 per process. This amounts to a saving of over $6,000 for six-point recorders that have no programmable control and no on-line accessibility to historical data. Nonetheless, we have at our disposal an indispensable tool in the computer that can optimize yields and reduce R&D time. To satisfy the metabolic demands of microbial cultures, computer control strategies can be employed for parameters, such as nutrient feeding, dissolved oxygen, and oxygen supplementation. Without requiring algorithms or equation-writing skills, operating conditions can be changed as a function of time or of any measured or calculated variable, such as cell density, protein concentration, and dissolved oxygen. With a few clicks of the mouse an investigator can change a digital readout of dissolved oxygen to an X-Y plot of that parameter, displaying a complete record of dissolved oxygen over time. With two more clicks of the mouse, X-Y plots of protein production and agitation can be superimposed over dissolved oxygen in a single chart to elucidate the interaction of all three parameters. Because of the proprietary nature of their work, some laboratories may be reluctant to consult with software manufacturers about manipulating the computer for yield improvement. However, many researchers who do use the computer to control the fermentation process have made significant advances in process development. In a feasibility study conducted at New Brunswick Scientific Co., Inc., researchers used a computer to design a nutrient feeding program and dissolved oxygen strategy for a fed-batch fermentation with continuous perfusion. Cell yields increased from 40 to 200 g/L (dry weight), and protein concentrations increased from 30 to 300 mg/L (3). If biochemical engineers are to make important progress in scientific research the computer must do more than store, retrieve, and display data. Scientists need to explore the many paths to increased cell growth and protein production. Instrument makers and the research community must cooperate more closely if they are to use the great potential of cybernetics in scientific research (Figure 7).

Figure 7. With currently available software programs, operators can produce a minute-byminute panorama of the complete process, observing on-line graphic displays of up to 24 parameters as well as interactions between variables.

19

Know Who Will Service Your System and Where

20

Ask About Start-Up Assistance and Service Agreements

Learn from the Mistakes of Others

Find out who will service your new equipment and where the service center is located. If service is not performed on site, equipment may need to be crated for return shipment to the factory. This is time-consuming and risky in terms of properly packaging equipment for transit. If service is performed at the vendors repair shop, then determine who is responsible for the crating, shipping, and insurance costs. If the equipment is sold and serviced by a major laboratory supply house, chances are good that the company will be around to deliver the support you need. Find out whether the service center has a large inventory of spare parts and how long it will take for parts to reach you.

To ensure that your new bioprocessing system is properly installed and operates efficiently, you and your staff should receive start-up assistance and training at your facility. The procedures involved in sterilizing, preparing, calibrating, and installing sensors and other ancillary equipment and instrumentation are vital to the success of your project, and the cost of conducting those procedures should be included in the price of the equipment. If the equipment fails to meet performance specifications, this is the time to find out.

Instrument makers are not known for volunteering detailed information about the products they manufacture. Sometimes it takes determination to get to all the facts. Unlike automobile companies who drown their consumers with information, technical data published by equipment manufacturers is frequently sparse. In the field of biotechnology, product quality does not come under the scrutiny of consumer protection agencies or the critical voice of the media. In the absence of a well-informed purchasing authority to help make important choices, researchers must be prepared to question product performance as well as a manufacturers credibility. For buyers with limited knowledge and practical experience in purchasing fermentation equipment, this guide is of particular value. Buyers with an engineering background and experience with fermentation or cell culture may be better equipped to find the best products at the best price. But even those with the know-how and experience can benefit from the mistakes of others.
References (1) Y. Chen et al., Production of Plasminogen Activator InhibitorType-1 (PAI-1) in an E. coli Fermentation Process, paper presented to the Ninth Symposium of the Protein Society, Boston, 1995. (2) J. Voss, Cleaning and Cleaning Validation: A Biotechnology Perspective, J. Voss, Ed. (Parenteral Drug Association, Bethesda, MD, 1996), 148. (3) Y. Chen et al., High Protein Expression Fermentation of Recombinant Pichia pastoris in Fed-Batch Process, unpublished paper, New Brunswick Scientific Co., Inc., Edison, NJ, and University of California, San Diego, La Jolla, 1995. BP