Neuroscientist Jack Turman collaborates with developmental biologists at the CCMB to understand the complex biology behind feeding behavior.

Eating seems like the most natural activity in the world.  We seem to be born knowing how to do it.  But in reality, the task of eating is a complex combination of neuromuscular circuitry, digestive preparedness, and learned behavior.  Each event is inextricably intertwined with the others and the success of the entire activity depends on their coordination in a specific period of time, near birth. Should one of these elements be amiss, an infant can suffer from malnutrition and a general inability to thrive.

USC Department of Biokinesiology and Physical Therapy Assistant Professor, Dr. Jack Turman spent years working with infants that had difficulty feeding.  His early work as a physical therapist allowed him to appreciate that only limited aspects of the 'failure to thrive' condition were understood.  Clinical observations taught health care providers and neuroscientists alike that, "Suckling is a very experience dependent behavior, or circuit".  As Turman explains, "You have to use it or you lose it."  What the biology was behind the acquisition of the vital and complicated activities of suckling and chewing was decidedly less clear. 

Even more distressing than the lack of knowledge about oral motor behaviors was the fact that as long as an infant’s gastrointestinal tract was in working order, the feeding problems were considered trivial by most on the clinical team.  I'd ask questions about how a particular health problem or injury would affect the brain development around feeding.  But I never really got any good answers or it just wasn't known.  All of this motivated me to get my Ph.D. in neuroscience and to focus on the development of oral motor function, explained Turman. 

Turman spent his first years as a researcher charting the expression of important neurotransmitters and neurotransmitter receptors in brainstem neurons critical for the production of mastication.  But his research efforts were aided considerably by a fortuitous discovery he made while performing an academic maintenance activity, a Medline search.  While combing through new journal articles looking for ties to his own studies, he happened upon an article about a transgenic knock-out mouse for the Krox-20 gene.  The Krox-20 deficient mice died shortly after birth, ostensibly due to an inability to feed.  Turman had found exactly what he was searching for, a promising genetic model to study oral motor behavior. 

The major developmental defect in the Krox-20 mice occurs during the early formation of neuronal structures.  A specific, segregated population of cranial neuron precursors is defective in the mutants.  In wild-type mice, these precursor cell aggregations, or rhombomeres, are organized into adjacent blocks and are part of the developing hindbrain. The Krox-20 mutation specifically the cells that emigrate from rhombomeres 3 and 5.

Fortunately for Turman, nearly everyone who works on the Krox-20 gene, including the authors who developed the transgenic mouse, analyzes the Krox-20 gene from the molecular level to determine what genes are being turned on and off.  But as Turman explains, They weren’t looking at the outcome.  Certainly nobody was looking in the craniofacial region. But they were very open that this might be a model of oral motor dysfunction.  So Turman ran with it.

It was at this point that Turman sought to more fully develop his collaborative relationship with scientists from the CCMB.  Turman turned to CCMB Director, Charles Shuler to bring in expertise on craniofacial development.  "It's rare enough to have a neuroscientist who studies the trigeminal nerve and how it relates to suckling and mastication, but to have that in conjunction with a craniofacial program is exceptionally rare," explained Turman.  Shuler and Turman decided to take a team approach to the analysis of the Krox-20 mouse, drawing on resources from both their labs. Turman offered neuroscience knowledge and experimental approaches: Shuler, craniofacial muscle biology expertise and graduate student hands.

As they came to discover, the Krox-20 mutation was not fatal in utero.  Rather, about 50-70% of the null mutant pups died within the first 24 hours.  Turman and his Craniofacial Biology graduate students, Nasrin Bahari and Shampa De, knew from gross behavioral observations of the pups that the opportunity to feed was there.  The pups were, for some reason, incapable of suckling.

The challenge to Turman, Bahari and De, was to figure out just why the pups were unable to perform this essential activity.  We started at the simplest level, with the muscles, explained Turman, We sectioned the cranial tissues of the pups to see what the mutation did to craniofacial development.  That is the easiest place to start because of the clear segregation of the craniofacial muscles. And that gave us our big finding.

The Krox-20 mutant pups were unable to feed because they lacked their primary jaw opener muscles, the anterior digastric and mylohyoid muscles.  Without functional primary jaw openers, the pups were unable to perform the power stroke of suckling behavior.  Closer inspection allowed the scientists to demonstrate that the jaw opening muscles actually formed in utero.  By the final quarter of development however, the muscles had atrophied extensively.  At birth, explained Turman, They’re gone.  In terms of movement execution, the loss of the anterior digastric and mylohyoid was critical and it was the only gross craniofacial difference Turman and the students noted between the knock-out mice and normal mice.

Being a neuroscientist, Turman next set his sights on analyzing the innervation of the jaw opening muscles.  What we've found is that the Krox-20 mutation causes a dramatic reduction in the trigeminal motor neurons and the reduction occurs prior to neuromuscular junction formation, describes Turman.  The group speculates that the Krox-20 mutation somehow eliminates the motor neurons so that the physical connection to the muscle never occurs.  With nothing to stimulate contraction, the anterior digastric and mylohyoid muscles simply waste away.

As the group continued with more detailed studies on the development of the neuromuscular connections, they discovered a rationale for why the jaw opening muscles are selectively affected. The control center of the craniofacial trigeminal motor nerve, its motor nucleus, gives rise to several branches of  motor neurons. Each of these branches connects with specific muscles of the jaw and each branch is rooted in a topographically distinct part of the control center. Neurons that innervate the jaw-opening muscles arise from the bottom portion of the control center. "The whole trigeminal motor nucleus is derived from the embryonic rhombomere structures, 1, 2 and 3.  So it makes sense that only the digastric and the mylohyoid are affected because the Krox-20 mutation specifically shuts down development of rhombomeres 3 and 5, explained Turman.

The Krox-20 mice exhibit an additional craniofacial phenotype of interest to dentists and those who study oral motor behavior.  The mice also have a loss of primary mesencephalic trigeminal sensory neurons.  These neurons provide sensory information to the jaw closer muscle spindles and interestingly a number of them innervate the periodontal ligament (PDL).  Sensory innvervation of the PDL is critical for mastication because it moderates how much signal comes back from the central nervous system through the motor nerves and thus how forcefully the jaw muscles close.  As Turman explains, "What we need to do is perform studies on the animals that live to see what happens to their PDL and what impact the loss has.  What complicates this is that the PDL receives some innervation from primary sensory neurons in the trigeminal ganglia.  The trigeminal ganglia is supposedly not affected by this Krox-20 mutation although we are not completely certain of this." 

Turman's small group has had access to the homozygous Krox-20 mice for only a year now.   But even in this short period of time, the Krox-20 knock-out mice have proven to be a truly useful model for studying oral motor disruptions.  With this single mutation, recaps Turman, You affect the function of the jaw opener muscles, reduce the number of trigeminal motor neurons and reduce the number of primary sensory neurons.

Turman does not know whether the Krox-20 mutation affects the proliferation of the neuronal precursors in the rhombomere or affects the survival of those cells so that there are insufficient numbers to make a normal trigeminal nerve. Future studies he proposes to perform will address these questions.  "The exciting thing for us is that nobody has looked at this level at what the mutation does.  There's still a lot of basic information to clear up.  But it's approachable and will provide us with big insights.  I just need the extra hands to do it," says Turman.

A selection of recent publications:

• Turman J Jr, Chopiuk NB, Shuler CF (in press).  The Krox-20 null mutation differentially affects the development of masticatory muscles.  Dev Neurosci.

• Turman J Jr, Hiyama L, Castillo M, Chandler S (2001).  Expression of group I and   metabotropic glutamate receptors in trigeminal neurons during postnatal development. Dev Neurosci. 23(1): 41-54.

• Turman JE Jr, MacDonald AS, Pawl KE, Bringas P, Chandler SH (2000). AMPA receptor subunit expression in trigeminal neurons during postnatal development. J Comp Neurol. 427(1): 109-23.

• Turman JE Jr, Ajdari J, Chandler SH (1999). NMDA receptor NR1 and NR2A/B subunit expression in trigeminal neurons during early postnatal development. J Comp Neurol. 409(2):237-49.